CN118343118B - Vehicle driving control method and system considering pedestrian crossing safety awareness - Google Patents

Abstract

The invention belongs to the field of road vehicle control, and relates to a vehicle driving control method and system considering pedestrian crossing safety awareness, wherein the system comprises a vehicle data acquisition module, a vehicle-mounted communication module, a pedestrian data acquisition module, a road side communication module, a central control module, a data processing module, a pedestrian safety awareness calculation module, an accelerator control module, a brake control module and a voice control module; the data processing module is used for processing pedestrian and vehicle data and extracting a series of behavior parameters capable of reflecting pedestrian crossing safety consciousness; the pedestrian safety awareness calculating module is used for calling a corresponding safety awareness detecting model according to a pedestrian crossing scene, and calculating the safety awareness score and grade of pedestrians when crossing the street by using the behavior parameters and the traffic flow parameters of the pedestrians. The system can objectively evaluate the pedestrian crossing safety awareness, and pre-warn and control vehicles with the residual collision time smaller than a preset value with pedestrians nearby according to the safety awareness level, so that traffic accidents are reduced.

Description

A vehicle driving control method and system considering pedestrian crossing safety awareness

Technical Field

The present invention belongs to the field of road vehicle control, and in particular, relates to a vehicle driving control method and system that takes into account the safety awareness of pedestrians crossing the street.

Background Art

The main reasons for pedestrians to get into traffic accidents are: (1) lack of vigilance, such as using mobile phones, tablets and other electronic devices while walking distracted, not fully observing oncoming vehicles when crossing the street, etc.; (2) violation of traffic rules, such as running red lights, crossing the road at non-pedestrian crossings, etc. These behaviors reflect the pedestrians' weak safety awareness.

Pedestrian safety awareness refers to their ability to recognize and adapt to the surrounding traffic environment and take necessary measures to ensure their own safety when participating in traffic activities. This awareness is mainly reflected in the degree of compliance with traffic rules, alertness to the traffic environment, and response speed to potential risks. Therefore, by studying the behavioral characteristics of pedestrians in street crossing scenes, a method for detecting pedestrian safety awareness is proposed, which has important practical significance. This method can be used to obtain the safety awareness level of pedestrians when crossing the street. When pedestrians with poor safety awareness are identified, nearby vehicles are controlled to slow down in advance and be ready to avoid them at any time. Targeted traffic safety education and warnings are also provided to such pedestrians to achieve a virtuous circle, thereby effectively reducing the occurrence of traffic accidents.

There are relatively few methods for testing pedestrian safety awareness. Existing methods mainly evaluate pedestrian safety awareness through safety awareness measurement questionnaires. For example, Rong Jing of Beijing Jiaotong University described in "Research on the Relationship between Traffic Safety Awareness and Safety Behavior" that the study mainly constructed a pedestrian safety awareness measurement questionnaire from four aspects: safety knowledge, safety attitude, safety behavior orientation, and psychological quality, and then used questionnaire surveys to evaluate safety awareness. Although this method can provide a certain degree of safety awareness evaluation, it has obvious limitations: this method is highly dependent on the subjective judgment and self-report of the participants, and the data is susceptible to memory bias and social expectations, which will limit the accuracy and reliability of the safety awareness detection results. In addition, this method lacks objective observation of pedestrians' immediate reactions and decisions in actual traffic environments, and these directly observed behaviors can more intuitively and effectively reveal pedestrians' safety awareness.

At present, there are no reports on the control of vehicles for pedestrian safety awareness detection, and there are only documents on detecting driver safety awareness. For example: Patent 201711345175.6 discloses a driver safety awareness assessment method based on virtual driving and EEG detection. This method collects driver operation data in a virtual driving environment, such as steering wheel rotation data, brake data, and turn signal opening data, to analyze the driver's behavior. At the same time, an EEG monitoring device is used to collect the driver's EEG signals to evaluate his emotional state and determine the driving personality. Finally, the driver's safety awareness is comprehensively evaluated based on the driver's driving behavior and driving personality. Although this method can provide a multi-dimensional safety awareness assessment, it is not suitable for pedestrian safety awareness detection. First, the driving operation data involved in this patent is not suitable for describing pedestrians. Secondly, the measurement of EEG signals is extremely sensitive to the environment and is easily affected by environmental interference, affecting its accuracy. In addition, wearing EEG monitoring equipment will also affect the normal actions of pedestrians, posing a safety risk.

Summary of the invention

In view of the above-mentioned technical problems and defects, the purpose of the present invention is to provide a vehicle driving control method that takes into account the safety awareness of pedestrians crossing the street. The method objectively evaluates the safety awareness of pedestrians by directly observing their immediate reactions and decisions in the actual traffic environment, and provides early warning and control of nearby vehicles whose remaining collision time with pedestrians is less than a predetermined value according to their safety awareness level, thereby reducing the occurrence of traffic accidents.

To achieve the above object, the present invention adopts the following technical solution:

A vehicle driving control method considering pedestrian crossing safety awareness, the method comprising the following steps:

Step S1: data acquisition;

Step S1.1. Obtain traffic flow data, the traffic flow data includes traffic volume , Traffic flow speed , Headway ;

Step S1.2. Obtain vehicle data, including the speed of the vehicle , the vehicle's position ( ), the steering angle of the vehicle when turning , the vehicle's wheelbase , vehicle stability factor , length of vehicle and width ;

Step S1.3. Collect pedestrian data;

Millimeter-wave radars and high-definition cameras are installed on the roadside at pedestrian crossings to collect the location of pedestrians ( ),speed , pedestrian stride length and shoulder width and real-time street crossing video;

Step S2: data processing;

After obtaining traffic flow data, vehicle data and pedestrian data, the data is processed to extract a series of behavioral parameters that can reflect pedestrians' safety awareness when crossing the street. Among them, in the scene of pedestrians crossing the street at signalized intersections, the behavioral parameters that can reflect pedestrians' safety awareness when crossing the street include: the ratio of the time pedestrians look at the signal light to the time they cross the street; , the average distance of pedestrians from the center line of the crosswalk , the frequency of pedestrians turning their heads while crossing the street , Pedestrians’ perception-reaction time to conflicting vehicles , the remaining collision time between pedestrians and conflicting vehicles , the time when pedestrians choose to cross the street within the signal cycle ;

In the scenario of pedestrians crossing the street at an unsignalized intersection, the behavioral parameters that can reflect pedestrians' safety awareness when crossing the street include: the time the pedestrians stay at the gaze point during the crossing , the average distance of pedestrians from the center line of the crosswalk , the frequency of pedestrians turning their heads while crossing the street , Pedestrians’ perception-reaction time to conflicting vehicles , the remaining collision time between pedestrians and conflicting vehicles ;

In the scene of pedestrians crossing the street, the behavioral parameters that can reflect the pedestrians' safety awareness when crossing the street include: the time the pedestrians stay at the gaze point while crossing the street , the average distance of pedestrians from the center line of the crosswalk , the frequency of pedestrians turning their heads while crossing the street , Pedestrians’ perception-reaction time to conflicting vehicles , the remaining collision time between pedestrians and conflicting vehicles , pedestrians approach obstructions and plan to observe before crossing the street ;

Step S3: Input the parameters extracted in different scenarios into the corresponding scoring functions;

Among them, the score function of the ratio of the time pedestrians look at the traffic light to the time they cross the street is:

;

In the formula, The ratio of the time pedestrians spend looking at the traffic light to the time they spend crossing the street; The ideal ratio of the time pedestrians spend looking at the traffic light to the time they spend crossing the street; is the standard deviation of the ideal ratio;

The score function for the time point at which a pedestrian chooses to cross the street within the signal cycle is:

;

In the formula, To help pedestrians choose the time to cross the street within the signal cycle in a pedestrian crossing scenario at a signalized intersection; It’s time for the green light to start; The duration of the green light;

The scoring function of the frequency of head turns when pedestrians cross the street is:

;

;

Where h is the frequency of pedestrians’ head rotation when crossing the street; is the ideal head rotation frequency, For traffic volume; is the traffic flow speed; is the headway time; is an adjustment function used to adjust the ideal head turning frequency according to the traffic volume, traffic flow speed and headway; All are adjustment coefficients, which adjust the degree of influence of traffic volume, traffic flow speed and headway on the ideal head turning frequency of pedestrians;

The pedestrian's perception-reaction time score function to the conflicting vehicle is:

;

In the formula, The perception-reaction time of pedestrians to conflicting vehicles; is the adjustment coefficient, which is used to adjust the speed of score increase or decrease;

The remaining collision time score function between pedestrians and conflicting vehicles is:

;

In the formula, is the remaining collision time between the pedestrian and the straight-moving vehicle; The remaining collision time between the pedestrian and the turning vehicle; The ideal remaining collision time between pedestrians and conflicting vehicles;

The score function for the average distance of pedestrians from the center line of the crosswalk is:

;

In the formula, is the average distance of pedestrians from the center line of the crosswalk; is the maximum permissible distance;

The score function of the dwell time of the gaze point when pedestrians are crossing the street is:

;

In the formula, For pedestrians crossing the street The dwell time of a fixation point; The total number of gaze points of pedestrians while crossing the street; For the ideal minimum residence time, is the ideal maximum residence time, It is an indicator function, and its value is 1 when the condition in the brackets is met, otherwise it is 0;

The scoring function for pedestrians to be alert to obstructions when crossing the street is:

;

;

In the formula, Provide pedestrians with time to observe when approaching obstructions and plan before crossing the street; ideal observation time for pedestrians approaching obstructions and planning before crossing the street; is the adjustment coefficient, which adjusts the speed at which the score decreases;

Step S4: using the established pedestrian crossing safety awareness detection model to detect pedestrian crossing safety awareness in different scenarios respectively; wherein the pedestrian crossing safety awareness detection model includes a detection model of pedestrian crossing safety awareness at a signalized intersection, which is used to detect pedestrian crossing safety awareness in a signalized intersection pedestrian crossing scenario; a detection model of pedestrian crossing safety awareness at an unsignalized intersection, which is used to detect pedestrian crossing safety awareness in an unsignalized intersection pedestrian crossing scenario; a detection model of pedestrian crossing safety awareness at a road section, which is used to detect pedestrian crossing safety awareness in a road section pedestrian crossing scenario;

The detection model of pedestrian crossing safety awareness at signalized intersections is as follows:

;

In the formula, Scoring pedestrians’ safety awareness when crossing a signalized intersection; All are weight coefficients;

Detection model of pedestrian safety awareness at unsignalized intersections:

;

In the formula, Scoring pedestrians’ safety awareness when crossing unsignalized intersections; All are weight coefficients;

Detection model of pedestrian safety awareness on road sections:

;

In the formula, : Safety awareness score of pedestrians when crossing the street; All are weight coefficients;

Step S5: grading the pedestrians' safety awareness according to their scores of safety awareness when crossing the street;

Step S6: Control nearby vehicles whose remaining collision time with pedestrians is less than a predetermined value according to the pedestrian's safety awareness level.

As a preferred embodiment of the present invention, the behavior parameters that can reflect the pedestrian's safety awareness when crossing the street in step S2 are extracted multiple times according to the detection interval during the pedestrian crossing process or are reused after the first extraction; wherein the pedestrian chooses the time point of crossing the street within the signal cycle Approach obstructions with pedestrians and plan observation time before crossing the street It is extracted during the first testing cycle, and the first extraction result is used for each subsequent test.

As a preferred embodiment of the present invention, in step S2, the pedestrian's perception-reaction time to the conflicting vehicle is It is used to measure the time it takes for pedestrians to perceive and respond to conflicting vehicles. This parameter cannot be extracted when there is no vehicle in conflict with pedestrians and takes a value of 0 until the pedestrian's perception-reaction time to the conflicting vehicle is extracted for the first time. In subsequent detections, if there is a new extraction result, it is used; otherwise, the average value of the previous extraction results is used. The remaining collision time between pedestrians and conflicting vehicles is used to reflect whether the pedestrian's position or action is safe. When there is no vehicle in conflict with pedestrians, the remaining collision time between pedestrians and conflicting vehicles takes a value of ∞ until the remaining collision time between pedestrians and conflicting vehicles is extracted for the first time. In subsequent detections, if there is a new extraction result, it is used; otherwise, the average value of the previous extraction results is used.

As a preferred embodiment of the present invention, in step S2, the remaining collision time between the pedestrian and the conflicting vehicle is The extraction method is: input the obtained pedestrian and vehicle data into the pedestrian crossing Model, the pedestrian crossing The model includes straight-moving vehicles and pedestrians Model, turning vehicles and pedestrians Model; in which straight-moving vehicles and pedestrians The model is as follows:

;

;

;

;

;

in, is the position of the vehicle; for the location of pedestrians; is the longitudinal distance between the vehicle and the pedestrian at the initial moment; is the lateral distance between the vehicle and the pedestrian at the initial moment; is the longitudinal distance of the conflict point offset; The distance from the front bumper of the vehicle to the offset conflict point; and are the length and width of the vehicle respectively; and are the pedestrian’s step length and shoulder width respectively; is the speed of the vehicle; is the speed of pedestrians; is the lateral speed of the pedestrian; is the longitudinal speed of the pedestrian;

Turning vehicles and pedestrians The model is as follows:

;

;

;

;

;

;

in, is the horizontal coordinate of the offset conflict point; is the position of the vehicle; for the location of pedestrians; and are the length and width of the vehicle respectively; and are pedestrian step length and shoulder width respectively; is the speed of the vehicle; is the speed of pedestrians; is the lateral speed of the pedestrian; is the longitudinal speed of the pedestrian; is the center angle of the turning vehicle’s front bumper at the initial collision point at the initial moment; The center angle of the circle from the front bumper of the turning vehicle to the offset conflict point; is the longitudinal distance from the pedestrian to the initial conflict point at the initial moment; is the longitudinal distance of the conflict point offset; is the longitudinal distance from the pedestrian to the offset conflict point; is the center angle The corresponding arc distance; is the center angle The corresponding arc distance; is the turning radius of the vehicle.

As a preferred embodiment of the present invention, in step S4, the detection model of pedestrian crossing safety awareness at the signalized intersection They are ; Detection model of pedestrian safety awareness at unsignalized intersections They are ; Detection model of pedestrian safety awareness when crossing the street They are .

As a preferred embodiment of the present invention, in step S5, 0-59 points are C level, 60-79 points are B level, and 80-100 points are A level. If the pedestrian's safety awareness level is A, the driver is reminded by voice that "there are pedestrians crossing the street ahead, please pay attention to avoid it"; if the pedestrian's safety awareness level is B, the driver is reminded by voice that "there are pedestrians crossing the street ahead, please pay attention to avoid it", and the throttle opening is reduced to the original If the pedestrian's safety awareness level is C, the driver will be reminded by voice that "there are pedestrians crossing the street ahead, please pay attention to avoid them" and the throttle opening will be reduced to the original , while improving the electronic hydraulic braking system to provide electronic braking force 1.5 times the original.

The present invention also provides a vehicle driving control system that takes into account pedestrian safety awareness when crossing the street in order to implement the above method. The system includes a vehicle data acquisition module, an on-board communication module, a pedestrian data acquisition module, a roadside communication module, a central control module, a data processing module, a pedestrian safety awareness calculation module, a throttle control module, a brake control module, and a voice control module;

Wherein, the vehicle data acquisition module is used to obtain the location and driving status information of the vehicle;

The vehicle communication module is used to upload the vehicle's driving status information and technical parameters to the central control module, and receive the control strategy issued by the central control module;

The pedestrian data collection module is used to obtain the location, speed and real-time street crossing video of pedestrians;

The roadside communication module is used to upload pedestrian data to the central control module;

The central control module is used to receive data uploaded by the vehicle communication module and the roadside communication module, obtain real-time traffic flow data from the traffic management department, and issue control strategies to vehicles according to the safety awareness level of pedestrians; it includes a data processing module and a pedestrian safety awareness calculation module;

The data processing module is used to process pedestrian and vehicle data and extract a series of behavioral parameters that can reflect pedestrian safety awareness when crossing the street, including the ratio of the time pedestrians stare at the traffic light to the time they cross the street. , the average distance of pedestrians from the center line of the crosswalk , the frequency of pedestrians turning their heads while crossing the street , Pedestrians’ perception-reaction time to conflicting vehicles , the remaining collision time between pedestrians and conflicting vehicles , the time when pedestrians choose to cross the street within the signal cycle , the time pedestrians spend looking at a point while crossing the street , pedestrians approach obstructions and plan to observe before crossing the street ;

The pedestrian safety awareness calculation module is used to call the corresponding safety awareness detection model according to the pedestrian crossing scene, and use the pedestrian's behavior parameters and traffic flow parameters to calculate the pedestrian's safety awareness score and level when crossing the street;

Among them, the detection model of pedestrian crossing safety awareness at signalized intersections is as follows:

;

;

;

;

;

;

;

;

In the formula, The ratio of the time pedestrians spend looking at the traffic light to the time they spend crossing the street; The ideal ratio of the time pedestrians spend looking at the traffic light to the time they spend crossing the street; is the standard deviation of the ideal ratio; is the average distance of pedestrians from the center line of the crosswalk; is the maximum permissible distance; To help pedestrians choose the time to cross the street within the signal cycle in a pedestrian crossing scenario at a signalized intersection; It’s time for the green light to start; is the duration of the green light; h is the frequency of the pedestrian's head turning when crossing the street; is the ideal head rotation frequency; For traffic volume; is the traffic flow speed; is the headway time; is an adjustment function used to adjust the ideal head turning frequency according to the traffic volume, traffic flow speed and headway; All are adjustment coefficients, which adjust the degree of influence of traffic volume, traffic flow speed and headway on the ideal head turning frequency of pedestrians; The perception-reaction time of pedestrians to conflicting vehicles; is the adjustment coefficient, which is used to adjust the speed of score increase or decrease; is the remaining collision time between the pedestrian and the straight-moving vehicle; The remaining collision time between the pedestrian and the turning vehicle; The ideal remaining collision time between pedestrians and conflicting vehicles; Scoring pedestrians’ safety awareness when crossing a signalized intersection; All are weight coefficients;

Detection model of pedestrian safety awareness at unsignalized intersections:

;

;

In the formula, For pedestrians crossing the street The dwell time of a fixation point; The total number of gaze points of pedestrians while crossing the street; is the ideal minimum stay time; is the ideal maximum residence time; It is an indicator function, and its value is 1 when the condition in the brackets is met, otherwise it is 0; Scoring pedestrians’ safety awareness when crossing unsignalized intersections; All are weight coefficients;

Detection model of pedestrian safety awareness on road sections:

;

;

;

In the formula, Provide pedestrians with time to observe when approaching obstructions and plan before crossing the street; ideal observation time for pedestrians approaching obstructions and planning before crossing the street; is the adjustment coefficient, which adjusts the speed at which the score decreases; Score pedestrians’ safety awareness when crossing the street; All are weight coefficients;

The throttle control module is used to control the throttle opening;

The brake control module is used to control the electronic power assist provided by the electronic hydraulic brake system;

The voice control module is used to control the voice system of the vehicle.

As a preferred embodiment of the present invention, the vehicle data acquisition module is connected to the vehicle-mounted radar camera to obtain vehicle driving state information, upload the vehicle driving state information and technical parameters to the central control module through the vehicle-mounted communication module, and store the data in the first memory;

The pedestrian data acquisition module is connected to the millimeter wave radar and high-definition camera set on the roadside to obtain the location, speed and real-time street crossing video of pedestrians, upload the pedestrian data to the central control module through the roadside communication module, and store the data in the second storage;

The central control module obtains real-time traffic flow data from the traffic management department and stores it in the third memory;

The data processing module retrieves vehicle data and pedestrian data from the first memory and the second memory, and uses a data processing method to extract a series of behavioral parameters that can reflect pedestrians' safety awareness when crossing the street;

The pedestrian safety awareness calculation module calls the corresponding safety awareness detection model according to the pedestrian crossing scene, and then uses the behavior parameters extracted by the data processing module and the traffic flow data retrieved from the third memory to calculate the pedestrian's safety awareness score and level, and stores the detection result in the fourth memory;

The central control module retrieves the pedestrian's safety awareness level information from the fourth memory and issues a vehicle control strategy according to the pedestrian's safety awareness level.

Advantages and beneficial effects of the present invention:

(1) The present invention directly observes the immediate reactions and decisions of pedestrians in actual traffic environments and objectively evaluates their safety awareness. This can not only provide early warning and control of nearby vehicles, but also provide targeted traffic safety education and warnings to pedestrians with low safety awareness. This method has very important practical significance for preventing and reducing the occurrence of traffic accidents.

(2) The method provided by the present invention is to get rid of the limitation of applicable scenarios and avoid using physiological parameters that are difficult to collect and easily affected by the environment. Instead, it adopts a series of behavioral parameters that can effectively reflect pedestrian safety awareness, such as the ratio of the length of time pedestrians stare at traffic lights to the length of time they cross the street, the time when pedestrians choose to cross the street within the signal cycle, the frequency of head turns when pedestrians cross the street, etc. These behavioral parameters can be directly observed in actual traffic scenarios, are easy to collect and are not easily affected by the environment, making the detection of pedestrian safety awareness when crossing the street more convenient for widespread application.

(3) The method provided by the present invention is effective in calculating the distance between pedestrians and vehicles. When considering the possibility that pedestrians may cross the street in a diagonal path, a straight-moving vehicle and pedestrian, a turning vehicle and pedestrian model, and then based on A detection model for pedestrian crossing safety awareness was established together with other pedestrian behavioral parameters. This method can improve the accuracy and reliability of pedestrian crossing safety awareness detection results, thereby better providing early warning and control for nearby vehicles.

BRIEF DESCRIPTION OF THE DRAWINGS

By referring to the following description in conjunction with the accompanying drawings, and with a more comprehensive understanding of the present invention, other objects and results of the present invention will become more apparent and easy to understand. In the accompanying drawings:

FIG1 is a flow chart of a vehicle driving control method considering pedestrian crossing safety awareness provided by Embodiment 1 of the present invention;

FIG2 is a schematic diagram of the TTC of a straight-moving vehicle and a pedestrian crossing the street according to the present invention;

FIG3 is a schematic diagram of TTC of a turning vehicle and a pedestrian crossing the street according to the present invention;

FIG4 is a structural block diagram of a vehicle driving control system that takes into account pedestrian crossing safety awareness, provided in Embodiment 2 of the present invention.

DETAILED DESCRIPTION

In order to enable those skilled in the art to better understand the technical solution and advantages of the present invention, the present application is described in detail below in conjunction with the accompanying drawings, but it is not intended to limit the protection scope of the present invention.

Embodiment 1 A vehicle driving control method considering pedestrian crossing safety awareness

The present embodiment provides a vehicle driving control method that takes into account the safety awareness of pedestrians crossing the street. FIG1 is a flow chart of a vehicle driving control method that takes into account the safety awareness of pedestrians crossing the street provided by the present embodiment.

As shown in FIG1 , a vehicle driving control method considering pedestrian crossing safety awareness provided in this embodiment includes the following steps:

Step S1: Data acquisition

Step S1.1. Obtain traffic flow data, the traffic flow data includes traffic volume , Traffic flow speed , Headway , can obtain real-time traffic flow data from the traffic management department;

Step S1.2. Obtain vehicle data, including the speed of the vehicle , the vehicle's position ( ), the steering angle of the vehicle when turning , the vehicle's wheelbase , vehicle stability factor , length of vehicle and width , the vehicle’s location, driving status and other data can be obtained through the vehicle’s networking facilities.

Step S1.3. Collect pedestrian data

Millimeter-wave radars, high-definition cameras and other detectors are installed on the side of the road at pedestrian crossings to collect the location of pedestrians ( ),speed , pedestrian stride length and shoulder width And real-time street crossing video and other data.

This embodiment statistically analyzes traffic accidents that occur when pedestrians are crossing the street, and divides typical scenes of pedestrian crossing into three categories: pedestrian crossing at signalized intersections, pedestrian crossing at unsignalized intersections, and pedestrian crossing on road sections.

In view of the situation where multiple pedestrians cross the street at the same time in the above scenario, this embodiment adopts measures such as pedestrian recognition, data synchronization and parallel processing to obtain the above data.

Specifically, at the moment a pedestrian enters the field of view of a high-definition camera, a unique identifier (ID) is assigned to each pedestrian and tracked continuously throughout the entire crossing process; then, a data fusion method is used to ensure the temporal synchronization and precise association of millimeter-wave radar data and video data, thereby accurately matching the position, speed and behavior data of each pedestrian; finally, multi-threading technology is used to process the data of each pedestrian in parallel.

Step S2: Data processing

After obtaining traffic flow data, vehicle data and pedestrian data, the data is processed to extract a series of behavioral parameters that can reflect pedestrians' safety awareness when crossing the street. Among them, in the scene of pedestrians crossing the street at signalized intersections, the behavioral parameters that can reflect pedestrians' safety awareness when crossing the street include: the ratio of the time pedestrians look at the signal light to the time they cross the street; , the average distance of pedestrians from the center line of the crosswalk , the frequency of pedestrians turning their heads while crossing the street , Pedestrians’ perception-reaction time to conflicting vehicles , the remaining collision time between pedestrians and conflicting vehicles , the time when pedestrians choose to cross the street within the signal cycle ;

In the scenario of pedestrians crossing the street at an unsignalized intersection, the behavioral parameters that can reflect pedestrians' safety awareness when crossing the street include: the time the pedestrians stay at the gaze point during the crossing , the average distance of pedestrians from the center line of the crosswalk , the frequency of pedestrians turning their heads while crossing the street , Pedestrians’ perception-reaction time to conflicting vehicles , the remaining collision time between pedestrians and conflicting vehicles ;

In the scene of pedestrians crossing the street, the behavioral parameters that can reflect the pedestrians' safety awareness when crossing the street include: the time the pedestrians stay at the gaze point while crossing the street , the average distance of pedestrians from the center line of the crosswalk , the frequency of pedestrians turning their heads while crossing the street , Pedestrians’ perception-reaction time to conflicting vehicles , the remaining collision time between pedestrians and conflicting vehicles , pedestrians approach obstructions and plan to observe before crossing the street .

In this embodiment, in order to timely warn and control nearby vehicles, the safety awareness of pedestrians when crossing the street needs to be detected multiple times according to a preset detection interval (determined by the crossing time, for example, a crossing time of 30 seconds can be detected once every 10 seconds, and a crossing time of 20 seconds can be detected once every 6 seconds). The above parameters are extracted multiple times according to the detection interval during the pedestrian crossing process or reused after the first extraction.

Specifically, the ratio of the time the pedestrian spends looking at the traffic light to the time the pedestrian spends crossing the street is : It is used to reflect the degree of attention paid by pedestrians to traffic lights when crossing a signalized intersection. If the ratio is too small, it means that pedestrians have poor safety awareness when crossing the street within the green light time. If the ratio is too large, it means that pedestrians have poor safety awareness in being alert to the surrounding traffic environment when crossing the street. An appropriate ratio means that pedestrians can cross the street while ensuring their own safety and complying with traffic lights, and have good safety awareness.

Extraction method: Use a visual saliency model (such as DeepGaze) to monitor and analyze the pedestrian's gaze focus during the crossing process, and identify the pedestrian's gaze time t ₁ on the traffic light during the detection cycle; then, compare this time with the detection cycle time t to extract the ratio of the pedestrian's gaze time on the traffic light to the crossing time, that is: .

The pedestrian chooses a time point to cross the street within the signal cycle : It is used to reflect the timing when pedestrians choose to cross the street within the traffic light cycle. Choosing to cross the street at the beginning of the green light is usually considered to be a manifestation of good safety awareness, while running a red light represents poor safety awareness.

Extraction method: Get the real-time status information of the intersection signal from the signal control system; then use the pedestrian position data provided by the millimeter-wave radar to monitor the time when the pedestrian starts to enter the crosswalk and record this time point; finally, compare the time when the pedestrian starts to cross the street with the signal cycle to extract the time when the pedestrian chooses to cross the street within the signal cycle. This parameter is extracted in the first detection cycle, and each subsequent detection uses the first extraction result.

The frequency of the pedestrian's head turning when crossing the street : Used to reflect pedestrians' alertness and attention to the dynamics of surrounding vehicles when crossing the street. Frequent head turns usually indicate better safety awareness.

Extraction method: Use posture estimation technology (such as OpenPose) to analyze the position and movement of pedestrians' heads in the video in real time, accurately track the direction changes of pedestrians' heads, and calculate the number of times the head turns within the detection cycle to extract the frequency of head turns when pedestrians cross the street.

in, is the detection cycle duration, is the number of head rotations during the detection cycle.

The pedestrian's perception-reaction time to the conflicting vehicle : A measure of the time it takes pedestrians to perceive and react to conflicting vehicles. Shorter perception-reaction times generally mean better vigilance.

Extraction method: Use object recognition and tracking algorithms (such as YOLO) to monitor the dynamic interaction between pedestrians and oncoming vehicles, and analyze the time interval from when pedestrians perceive the conflicting vehicle to when they take evasive action. This time interval is the perception-reaction time of pedestrians to the conflicting vehicle. This parameter cannot be extracted when there is no vehicle in conflict with pedestrians, and the value is 0 until the perception-reaction time of pedestrians to the conflicting vehicle is extracted for the first time. In subsequent detections, if there is a new extraction result, it will be used, otherwise the average value of the previous extraction results (non-0) will be used.

The remaining collision time between the pedestrian and the conflicting vehicle (Time to Collision, ): If there is a conflict point with the pedestrian and the vehicle while maintaining the current speed and direction, this is the remaining time before the pedestrian and the vehicle collide. The longer it is, the safer the pedestrian's position or action is. The longer the remaining collision time between pedestrians and conflicting vehicles during the street crossing, the better the pedestrian's safety awareness is.

Extraction method: Input the acquired pedestrian and vehicle data into the pedestrian crossing The model extracts the remaining collision time between pedestrians and conflicting vehicles. When there is no vehicle in conflict with pedestrians, this parameter takes the value of ∞ until the remaining collision time between pedestrians and conflicting vehicles is extracted for the first time. In subsequent detections, if there is a new extraction result, it is used, otherwise the average value of the previous extraction results (non-∞) is used.

The average distance of pedestrians from the center line of the crosswalk : Measures the extent to which pedestrians follow street crossing regulations. Walking close to the center line of the crosswalk is generally seen as safer and more compliant with traffic rules, representing a better sense of safety.

Extraction method: Use target detection and image segmentation algorithms (such as Mask R-CNN) to identify and mark the precise location and boundaries of the crosswalk, track the path of pedestrians while crossing the street, and accurately calculate the average lateral distance of pedestrians relative to the center line of the crosswalk during the detection cycle, thereby extracting the average distance of pedestrians from the center line of the crosswalk.

in, is the number of video image frames within the detection period, For the The distance of the pedestrian from the center line of the crosswalk in the frame image.

The time the pedestrian spends looking at the point while crossing the street It refers to the time that pedestrians keep their gaze on any person or object when crossing the street. It reflects whether pedestrians can effectively allocate their attention to different potential risk sources when crossing unsignalized intersections and sections, and maintain comprehensive monitoring of the surrounding environment. The ideal gaze point dwell time allows pedestrians to obtain necessary information without missing other risk information, which represents better safety awareness.

Extraction method: Use a visual saliency model (such as DeepGaze) to monitor and analyze the gaze focus of pedestrians when crossing the street, determine the time the pedestrian's gaze stays at each gaze point during the detection cycle, and extract the time the pedestrian stays at the gaze point while crossing the street.

The pedestrian approaches the obstruction and plans to observe the time before crossing the street It means that when pedestrians cross the street, they may encounter vehicles parked on the side of the road blocking their view. When the straight-line distance between the pedestrian and the obstruction is less than 1 meter, the pedestrian should slow down or stop moving and observe the oncoming vehicles. The longer the observation lasts, the better the pedestrian's safety awareness.

Extraction method: Use object recognition and tracking algorithms (such as YOLO) to accurately identify and mark potential line-of-sight obstructions such as parked vehicles in the video, and work with millimeter-wave radar to track the behavior of pedestrians when they approach these obstructions. When the straight-line distance between the pedestrian and the obstruction is less than 1 meter, the pedestrian is determined to be approaching the obstruction at this time. The observation time before planning to cross the street is then counted from when the pedestrian slows down or stops moving until he starts to cross the street. The observation time before the pedestrian approaches the obstruction and plans to cross the street is extracted. This parameter is extracted in the first detection cycle, and the first extraction result is used for each subsequent detection.

It should be noted that the method for extracting the above parameters provided in this embodiment is for reference only. Those skilled in the art may also adopt other methods to obtain the above parameters. This application does not impose any limitation on the method for extracting the above parameters.

In this embodiment, in order to accurately obtain the remaining collision time between the pedestrian and the conflicting vehicle, , establish pedestrian crossing Model, the pedestrian crossing The model includes straight-moving vehicles and pedestrians Model, turning vehicles and pedestrians Model.

In reality, pedestrians’ paths when crossing the street are not always a straight line perpendicular to the road. Therefore, the pedestrian’s path may be a diagonal line. When the pedestrian’s path is a diagonal line, the longitudinal speed of the pedestrian will cause the original conflict point to shift (as shown in FIG. 2 ). Based on this situation, the collision path between the straight-moving vehicle and the pedestrian established in this embodiment is The model is as follows:

;

;

;

;

;

in, is the position of the vehicle; for the location of pedestrians; is the longitudinal distance between the vehicle and the pedestrian at the initial moment; is the lateral distance between the vehicle and the pedestrian at the initial moment; is the longitudinal distance of the conflict point offset; The distance from the front bumper of the vehicle to the offset conflict point; and are the length and width of the vehicle respectively; and are the pedestrian’s step length and shoulder width respectively; is the speed of the vehicle; is the speed of pedestrians; is the lateral speed of the pedestrian; is the longitudinal speed of the pedestrian;

When the pedestrian's crossing path is a straight line, the conflict point is fixed. , , , , When it is infinite, it means there is no collision risk between the vehicle and the pedestrian.

This embodiment also establishes the Model, assuming that the vehicle has completed the deceleration operation before turning and maintains a constant speed during the turning process. If the pedestrian's crossing path is a diagonal line, the pedestrian's longitudinal speed will also cause the original conflict point to shift (as shown in Figure 3). At this time, the horizontal coordinate of the conflict point between the pedestrian and the turning vehicle needs to be calculated. ,according to Equi-parameter calculation ;

The horizontal coordinate of the conflict point between pedestrians and turning vehicles The calculation formula is as follows:

;

;

;

;

;

;

in, for pedestrian trajectories; is the vehicle trajectory; is the turning radius of the vehicle; is the stability factor of the vehicle; is the vehicle wheelbase; is the speed of the vehicle; The steering angle of the vehicle is the angle formed by the center line of the vehicle when the front wheels turn left or right and the center line when the front wheels do not deflect; is the lateral speed of the pedestrian; is the longitudinal speed of the pedestrian; for the location of pedestrians; The specific value of is determined by the pedestrian's movement direction.

Turning vehicles and pedestrians The model is as follows:

;

;

;

;

;

;

in, is the horizontal coordinate of the offset conflict point; is the position of the vehicle; for the location of pedestrians; and are the length and width of the vehicle respectively; and are pedestrian step length and shoulder width respectively; is the speed of the vehicle; is the speed of pedestrians; is the lateral speed of the pedestrian; is the longitudinal speed of the pedestrian; is the center angle of the turning vehicle’s front bumper at the initial collision point at the initial moment; The center angle of the circle from the front bumper of the turning vehicle to the offset conflict point; is the longitudinal distance from the pedestrian to the initial conflict point at the initial moment; is the longitudinal distance of the conflict point offset; is the longitudinal distance from the pedestrian to the offset conflict point; is the center angle The corresponding arc distance; is the center angle The corresponding arc distance; is the turning radius of the vehicle.

When the pedestrian's crossing path is a straight line, the conflict point is fixed. , , , , , When it is infinite, it means there is no collision risk between the vehicle and the pedestrian.

Step S3: Input the parameters extracted in different scenarios into the corresponding scoring functions;

Among them, the score function of the ratio of the time pedestrians look at the traffic light to the time they cross the street is:

;

in, The ratio of the time pedestrians spend looking at the traffic light to the time they spend crossing the street; is the ideal ratio of the time pedestrians spend looking at the traffic light to the time they spend crossing the street. ; is the standard deviation of the ideal ratio, .

In this embodiment, Measures the ratio of the time pedestrians pay attention to traffic lights at signalized intersections to the time they spend crossing the street. The closer this ratio is to the ideal ratio , indicating that pedestrians have a good sense of compliance with signals and safety, and deviate from The score for this item will be reduced.

The score function for the time point at which a pedestrian chooses to cross the street within the signal cycle:

;

in, To help pedestrians choose the time to cross the street within the signal cycle in a pedestrian crossing scenario at a signalized intersection; It’s time for the green light to start; The length of the green light. An earlier choice usually indicates a higher safety awareness. Choosing to cross the street at the end of the green light will reduce the score. Choosing to cross the street during the yellow or red light indicates a poor safety awareness and requires a penalty to reduce the score.

Score function for the frequency of head turns when pedestrians cross the street:

;

;

Where h is the frequency of pedestrians’ head rotation when crossing the street; is the ideal head rotation frequency, ; For traffic volume; is the traffic flow speed; is the headway time; is an adjustment function used to adjust the ideal head turning frequency according to the traffic volume, traffic flow speed and headway; They are adjustment coefficients, which adjust the influence of traffic volume, traffic flow speed and headway on the ideal head turning frequency of pedestrians. to They are ;

In this embodiment, h reflects the pedestrian's alertness. When the head rotation frequency is greater than the ideal frequency, the score is 1, and when the head rotation frequency is less than the ideal frequency, the score decreases. The head rotation frequency of pedestrians crossing the street is easily affected by traffic flow, so in order to ensure the fairness of the pedestrian's score, it is necessary to adjust the ideal head rotation frequency according to different traffic flow conditions.

Pedestrian perception of conflicting vehicles-reaction time score function:

;

in, The perception-reaction time of pedestrians to conflicting vehicles; is the adjustment coefficient, which is used to adjust the speed of score increase or decrease. The value range is .

In this embodiment, It indicates the time it takes pedestrians to perceive and respond to conflicting vehicles. A shorter perception-reaction time indicates higher alertness and better safety awareness. The shorter the perception-reaction time, the higher the score.

The remaining collision time score function between pedestrians and conflicting vehicles:

;

in, is the remaining collision time between the pedestrian and the straight-moving vehicle; The remaining collision time between the pedestrian and the turning vehicle; is the ideal remaining collision time between the pedestrian and the conflicting vehicle, .

In this embodiment, Predict the remaining time before a pedestrian and a vehicle collide at the current speed and direction. Indicates a lower collision risk. The hyperbolic tangent function is used to ensure that the function value is within a reasonable range. The function input selects the minimum remaining collision time between the pedestrian and the conflicting vehicle within the detection cycle. When the value is greater than hour, The function is close to 1, the score is high, and when the collision time is less than hour, The function quickly approaches 0 and the score drops rapidly.

The average distance of pedestrians from the center line of the crosswalk score function:

;

in, is the average distance of pedestrians from the center line of the crosswalk; is the maximum allowable distance, The value is half of the length of the crosswalk.

In this embodiment, Reflects the walking position of pedestrians on the crosswalk. The closer to the center line, the safer it is. is the maximum tolerated deviation distance. If this distance is exceeded, the score drops to 0, indicating that the pedestrian has deviated from the safe crossing area.

The score function of the dwell time of the gaze point when pedestrians cross the street:

;

in, For pedestrians crossing the street The dwell time of a fixation point; The total number of gaze points of pedestrians while crossing the street; For the ideal minimum residence time, ; is the ideal maximum residence time, ; It is an indicator function. When the condition in the brackets is met, the value is 1, otherwise it is 0.

In this embodiment, is the time a pedestrian spends looking at a certain point when crossing the street. In the ideal range Insiders can obtain necessary information without missing other risk information and have a high sense of safety. Therefore, the greater the proportion of fixation points whose dwell time falls within the ideal range, the higher the score for this item.

The scoring function for pedestrians to be alert to occlusions when crossing the street:

;

;

in, Provide pedestrians with time to observe when approaching obstructions and plan before crossing the street; The ideal observation time for pedestrians approaching obstructions and planning before crossing the street, ; is the adjustment coefficient, which adjusts the speed at which the score decreases. Its value range is ;

Called when pedestrians cross the street in a section with obstructed vision The function calculates the score of pedestrians alert to occlusions. At least When the pedestrian fully observes the oncoming vehicle, he/she has a high safety awareness. Lower than , indicating that the pedestrian’s vigilance is poor and the score for this item is reduced; when a pedestrian crosses the street without any visual obstruction, the score for this item is 1.

Step S4: using the established pedestrian crossing safety awareness detection model to detect pedestrian crossing safety awareness in different scenarios respectively; wherein the pedestrian crossing safety awareness detection model includes a detection model of pedestrian crossing safety awareness at a signalized intersection, which is used to detect pedestrian crossing safety awareness in a signalized intersection pedestrian crossing scenario; a detection model of pedestrian crossing safety awareness at an unsignalized intersection, which is used to detect pedestrian crossing safety awareness in an unsignalized intersection pedestrian crossing scenario; a detection model of pedestrian crossing safety awareness at a road section, which is used to detect pedestrian crossing safety awareness in a road section pedestrian crossing scenario;

In this embodiment, according to the behavior parameters that can reflect the pedestrian's safety awareness when crossing the street at a signalized intersection (the ratio of the pedestrian's gaze time at the signal light to the crossing time), , the average distance of pedestrians from the center line of the crosswalk , the frequency of pedestrians turning their heads while crossing the street , Pedestrians’ perception-reaction time to conflicting vehicles , the remaining collision time between pedestrians and conflicting vehicles , the time when pedestrians choose to cross the street within the signal cycle ) The detection model of pedestrian crossing safety awareness at signalized intersections is established as follows:

;

in, Scoring pedestrians’ safety awareness when crossing a signalized intersection; are weight coefficients, Specifically, They are .

According to the behavioral parameters that can reflect pedestrians' safety awareness when crossing the street at an unsignalized intersection (the time the pedestrians stay at the gaze point during the crossing), , the average distance of pedestrians from the center line of the crosswalk , the frequency of pedestrians turning their heads while crossing the street, and the pedestrians’ perception-reaction time to conflicting vehicles , the remaining collision time between pedestrians and conflicting vehicles ) Establish a detection model for pedestrian safety awareness at unsignalized intersections:

;

in, Scoring pedestrians’ safety awareness when crossing unsignalized intersections; are weight coefficients, Specifically, They are .

According to the behavior parameters that can reflect the pedestrians' safety awareness when crossing the street (the time the pedestrians stay at the gaze point when crossing the street), , the average distance of pedestrians from the center line of the crosswalk , the frequency of pedestrians turning their heads while crossing the street, and the pedestrians’ perception-reaction time to conflicting vehicles , the remaining collision time between pedestrians and conflicting vehicles , pedestrians approach obstructions and plan to observe before crossing the street ) Establish a detection model for pedestrian safety awareness when crossing the street:

;

in, : Safety awareness score of pedestrians when crossing the street; are weight coefficients, Specifically, They are .

Step S5: grading the pedestrians' safety awareness according to their scores of safety awareness when crossing the street;

Specifically, 0-59 points are C grade, 60-79 points are B grade, and 80-100 points are A grade; among them, pedestrians with a safety awareness grade of A have good safety awareness and are less likely to have a traffic accident when crossing the street; pedestrians with a safety awareness grade of B have poor safety awareness and are more likely to have a traffic accident when crossing the street; pedestrians with a safety awareness grade of C have very poor safety awareness and are very likely to have a traffic accident when crossing the street.

Step S6: Control nearby vehicles whose remaining collision time with pedestrians is less than a predetermined value (eg, 2.5 s) according to the pedestrian's safety awareness level.

In this embodiment, different control strategies are corresponding to different safety awareness levels. Specifically, if the pedestrian's safety awareness level is A, the driver is reminded by voice that "there are pedestrians crossing the street ahead, please pay attention to avoid it"; if the pedestrian's safety awareness level is B, the driver is reminded by voice that "there are pedestrians crossing the street ahead, please pay attention to avoid it", and the throttle opening is reduced to the original If the pedestrian's safety awareness level is C, the driver will be reminded by voice that "there is a pedestrian crossing the street ahead, please give way" and the throttle opening will be reduced to the original , while improving the electronic hydraulic braking system to provide electronic braking force 1.5 times the original.

In this embodiment, pedestrian warnings can also be issued based on the pedestrian's safety awareness level. After the pedestrian completes crossing the street, his final safety awareness level can be determined based on the average of multiple safety awareness scores. If the pedestrian's safety awareness level is A, no supervision is required; if the pedestrian's safety awareness level is B, the traffic management department needs to give corresponding warnings in the form of text messages or phone calls to urge pedestrians to improve their safety awareness when crossing the street; if the pedestrian's safety awareness level is C, the traffic management department needs to give corresponding warnings in the form of text messages or phone calls, and provide targeted traffic safety education to such pedestrians, to help such pedestrians learn traffic safety knowledge and improve their safety awareness when crossing the street.

Example 2 A vehicle driving control system considering pedestrian crossing safety awareness

As shown in FIG4 , a vehicle driving control system that considers pedestrian crossing safety awareness provided in this embodiment includes: a vehicle data acquisition module, an on-board communication module, a pedestrian data acquisition module, a roadside communication module, a central control module, a data processing module, a pedestrian safety awareness calculation module, a throttle control module, a brake control module, and a voice control module;

The vehicle data acquisition module is used to obtain the location and driving status information of the vehicle;

The vehicle communication module is used to upload the vehicle's driving status information and technical parameters to the central control module, and receive the control strategy issued by the central control module;

The pedestrian data collection module is used to obtain data such as the location, speed and real-time street crossing video of pedestrians;

The roadside communication module is used to upload pedestrian data to the central control module;

The central control module is used to receive data uploaded by the vehicle communication module and the roadside communication module, obtain real-time traffic flow data from the traffic management department, and issue control strategies to vehicles according to the safety awareness level of pedestrians; it includes a data processing module and a pedestrian safety awareness calculation module;

The data processing module is used to process pedestrian and vehicle data and extract a series of behavioral parameters that can reflect pedestrian safety awareness when crossing the street, including the ratio of the time pedestrians stare at the traffic light to the time they cross the street. , the average distance of pedestrians from the center line of the crosswalk , the frequency of pedestrians turning their heads while crossing the street , Pedestrians’ perception-reaction time to conflicting vehicles , the remaining collision time between pedestrians and conflicting vehicles , the time when pedestrians choose to cross the street within the signal cycle , the time pedestrians spend looking at a point while crossing the street , pedestrians approach obstructions and plan to observe before crossing the street ;

The pedestrian safety awareness calculation module is used to call the corresponding safety awareness detection model according to the pedestrian crossing scene, and use the pedestrian's behavior parameters and traffic flow parameters to calculate the pedestrian's safety awareness score and level when crossing the street;

Among them, the detection model of pedestrian crossing safety awareness at signalized intersections is as follows:

;

;

;

;

;

;

;

;

in, The ratio of the time pedestrians spend looking at the traffic light to the time they spend crossing the street; is the ideal ratio of the time pedestrians spend looking at the traffic light to the time they spend crossing the street. ; is the standard deviation of the ideal ratio; is the average distance of pedestrians from the center line of the crosswalk; is the maximum allowable distance, The value is half of the length of the crosswalk; Choose a time for pedestrians to cross the street within the signal cycle; It’s time for the green light to start; is the duration of the green light; h is the frequency of the pedestrian's head turning when crossing the street; is the ideal head rotation frequency, ; For traffic volume; is the traffic flow speed; is the headway time; is an adjustment function used to adjust the ideal head turning frequency according to the traffic volume, traffic flow speed and headway; All are adjustment coefficients, which adjust the degree of influence of traffic volume, traffic flow speed and headway on the ideal head turning frequency of pedestrians; The perception-reaction time of pedestrians to conflicting vehicles; is the adjustment coefficient, which is used to adjust the speed of score increase or decrease. The value range is ; is the remaining collision time between the pedestrian and the straight-moving vehicle; The remaining collision time between the pedestrian and the turning vehicle; is the ideal remaining collision time between the pedestrian and the conflicting vehicle, ; Scoring pedestrians’ safety awareness when crossing a signalized intersection; are weight coefficients, respectively ;

Detection model of pedestrian safety awareness at unsignalized intersections:

;

;

in, For pedestrians crossing the street The dwell time of a fixation point; The total number of gaze points of pedestrians while crossing the street; For the ideal minimum residence time, ; is the ideal maximum residence time, ; It is an indicator function, and its value is 1 when the condition in the brackets is met, otherwise it is 0; Scoring pedestrians’ safety awareness when crossing unsignalized intersections; are weight coefficients, respectively .

Detection model of pedestrian safety awareness on road sections:

;

;

;

in, Provide pedestrians with time to observe when approaching obstructions and plan before crossing the street; The ideal observation time for pedestrians approaching obstructions and planning before crossing the street, ; is the adjustment coefficient, which adjusts the speed at which the score decreases; Score pedestrians’ safety awareness when crossing the street; are weight coefficients, respectively .

The throttle control module is used to control the throttle opening;

The brake control module is used to control the electronic power assist provided by the electronic hydraulic brake system;

The voice control module is used to control the voice system of the vehicle.

In the specific implementation process, the vehicle data acquisition module is connected to the vehicle radar camera to obtain the vehicle driving status information, upload the vehicle driving status information and technical parameters to the central control module through the vehicle communication module, and store the data in the first memory;

The pedestrian data acquisition module is connected to the millimeter wave radar and high-definition camera set on the roadside to obtain data such as the location, speed and real-time street crossing video of pedestrians, upload the pedestrian data to the central control module through the roadside communication module, and store the data in the second storage;

The central control module obtains real-time traffic flow data from the traffic management department and stores it in the third memory;

The data processing module retrieves vehicle data and pedestrian data from the first memory and the second memory, and uses data processing technology and methods to extract a series of behavioral parameters that can reflect pedestrians' safety awareness when crossing the street;

The pedestrian safety awareness calculation module calls the corresponding safety awareness detection model according to the pedestrian crossing scene, and then uses the behavior parameters extracted by the data processing module and the traffic flow data retrieved from the third memory to calculate the pedestrian's safety awareness score and level, and stores the detection result in the fourth memory;

The central control module retrieves the pedestrian's safety awareness level information from the fourth memory, and issues a vehicle control strategy according to the pedestrian's safety awareness level:

(1) If the pedestrian’s safety awareness level is A, the driver will be reminded by voice: “There is a pedestrian crossing the street ahead, please give way.”

(2) If the pedestrian's safety awareness level is B, the driver will be reminded by voice that "there is a pedestrian crossing the street ahead, please give way" and the throttle opening will be reduced to the original .

(3) If the pedestrian's safety awareness level is C, the driver will be reminded by voice that "there is a pedestrian crossing the street ahead, please give way" and the throttle opening will be reduced to the original , while improving the electronic hydraulic braking system to provide electronic braking force 1.5 times the original.

The traffic management department periodically retrieves the pedestrian's safety awareness level information from the fourth memory and issues a warning based on the pedestrian's final safety awareness level:

(1) If the pedestrian’s safety awareness level is A, no supervision is required.

(2) If the pedestrian’s safety awareness level is B, a corresponding warning needs to be given in the form of text messages or phone calls to urge pedestrians to improve their safety awareness when crossing the street.

(3) If the pedestrian’s safety awareness level is C, a corresponding warning should be given in the form of text messages or phone calls, and targeted traffic safety education should be provided to help such pedestrians learn traffic safety knowledge and improve their safety awareness when crossing the street.

The present invention also provides an electronic device, comprising: one or more processors and a memory; wherein the memory is used to store one or more programs, and when the one or more programs are executed by the one or more processors, the one or more processors implement a vehicle driving control method that takes into account the safety awareness of pedestrians crossing the street as described in Example 1.

The present invention also provides a computer-readable medium having a computer program stored thereon, and when the computer program is executed by a processor, the vehicle driving control method taking into account the safety awareness of pedestrians crossing the street as described in Example 1 is implemented.

Those skilled in the art will appreciate that all or part of the functions of the various methods/modules in the above embodiments may be implemented by hardware or by computer programs. When all or part of the functions in the above embodiments are implemented by computer programs, the program may be stored in a computer-readable storage medium, which may include: a read-only memory, a random access memory, a disk, an optical disk, a hard disk, etc. The program is executed by a computer to implement the above functions. For example, the program is stored in the memory of the device, and when the program in the memory is executed by the processor, all or part of the above functions can be implemented.

In addition, when all or part of the functions in the above-mentioned embodiments are implemented by means of a computer program, the program can also be stored in a storage medium such as a server, another computer, a disk, an optical disk, a flash drive or a mobile hard disk, and saved to the memory of a local device by downloading or copying, or the system of the local device is updated. When the program in the memory is executed by the processor, all or part of the functions in the above-mentioned embodiments can be implemented.

The above specific examples are used to illustrate the present invention, which is only used to help understand the present invention and is not intended to limit the present invention. For those skilled in the art of the present invention, according to the idea of the present invention, some simple deductions, deformations or substitutions can be made. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims (8)

1. A vehicle driving control method considering pedestrian crossing safety awareness, characterized by comprising the steps of:

Step S1: obtaining data;

S1.1, acquiring traffic flow data, wherein the traffic flow data comprises traffic volume Traffic flow velocityTime distance of head of a vehicle；

Step S1.2. Acquiring vehicle data, the vehicle data comprising a speed of the vehiclePosition of vehicle) Steering angle when turning a vehicleWheelbase of vehicleStability factor of vehicleLength of vehicleSum width of；

S1.3, collecting pedestrian data;

millimeter wave radar and high-definition camera are arranged on the road side of pedestrian crossing and pedestrian position is collected ) Speed and velocity ofStep size of pedestrianAnd shoulder widthReal-time street crossing video;

Step S2: data processing;

After the traffic flow data, the vehicle data and the pedestrian data are acquired, the data are processed, and a series of behavior parameters capable of reflecting pedestrian crossing safety awareness are extracted; under the scene that the pedestrians cross the street at the signalized intersection, the behavior parameters capable of reflecting the pedestrian's street-crossing safety awareness include: pedestrian gazes time length to signal lamp and crosses street time length ratio Average distance of pedestrian from crosswalk centerlineHead rotation frequency in pedestrian crossing roadPerception-response time of pedestrian to conflicting vehiclesRemaining time of collision of pedestrian with conflicting vehicleThe time point of pedestrian crossing street in signal period；

Under the scene of no signalized intersection pedestrian crossing, the behavior parameters capable of reflecting pedestrian crossing safety consciousness include: fixation point stay time on pedestrian crossing roadAverage distance of pedestrian from crosswalk centerlineHead rotation frequency in pedestrian crossing roadPerception-response time of pedestrian to conflicting vehiclesRemaining time of collision of pedestrian with conflicting vehicle；

Under the road section pedestrian crossing scene, the behavior parameters capable of reflecting pedestrian crossing safety consciousness include: fixation point stay time on pedestrian crossing roadAverage distance of pedestrian from crosswalk centerlineHead rotation frequency in pedestrian crossing roadPerception-response time of pedestrian to conflicting vehiclesRemaining time of collision of pedestrian with conflicting vehiclePedestrian approaches the shelter and plans the time of observation before crossing the street；

Step S3: inputting parameters extracted in different scenes into a scoring function corresponding to the parameters;

the ratio scoring function of the gazing time length of the pedestrian to the signal lamp and the street crossing time length is as follows:

；

in the method, in the process of the invention, The fixation time length of the signal lamp is compared with the street crossing time length for pedestrians; the ratio of the gazing time length of the ideal pedestrian to the signal lamp to the street crossing time length is provided; standard deviation of ideal ratio;

the time point scoring function of pedestrian selection crossing street in the signal period is as follows:

；

in the method, in the process of the invention, Selecting a time point of crossing a street in a signal period for a pedestrian in a pedestrian crossing scene at a signal intersection; The starting time of the green light is the starting time of the green light; the green light time is the green light time;

the head rotation frequency scoring function in the pedestrian crossing course is:

；

；

wherein h is the head rotation frequency in the process of pedestrian crossing; For a desired head rotation frequency, Is the traffic volume; is the traffic flow speed; Is the time interval of the head; the head rotation frequency is used for adjusting the ideal head rotation frequency according to traffic volume, traffic flow speed and headway; the method is characterized in that the method is an adjusting coefficient, and the influence degree of traffic volume, traffic flow speed and headway on the ideal head rotation frequency of pedestrians is adjusted;

The pedestrian's perception-response time score function for conflicting vehicles is:

；

in the method, in the process of the invention, A perceived-reaction time for a pedestrian to a conflicting vehicle; for adjusting the coefficient, the speed of increasing and decreasing the fraction is adjusted;

the remaining collision time score function for pedestrians and conflicting vehicles is:

；

in the method, in the process of the invention, Remaining collision time for pedestrians and straight vehicles; Remaining collision time for pedestrians and turning vehicles; Ideal remaining collision time for pedestrians and conflicting vehicles;

the average distance score function of pedestrians from the center line of the crosswalk is as follows:

；

in the method, in the process of the invention, The average distance between the pedestrian and the center line of the crosswalk; Is the maximum allowable distance;

The gaze point dwell time scoring function on the way through the pedestrian is:

；

in the method, in the process of the invention, The pedestrian is on the way of crossing the streetDwell time of the individual gaze points; the total number of fixation points on the way of pedestrian crossing; For a desired minimum residence time, For a desired maximum residence time,To indicate a function, when the condition in brackets is satisfied, the value is 1, otherwise 0;

the scoring function of the alert shielding object when pedestrians cross the street is as follows:

；

；

in the method, in the process of the invention, The method comprises the steps of approaching a shelter for pedestrians and planning the observation time before crossing a street; an ideal observation time before the pedestrian approaches the shelter and plans to cross the street; to adjust the coefficient, adjusting the rate of fractional reduction;

Step S4: the pedestrian crossing safety consciousness detection model is used for respectively detecting pedestrian crossing safety consciousness under different scenes; the pedestrian crossing safety awareness detection model comprises a detection model of pedestrian crossing safety awareness of a signalized intersection, and is used for detecting pedestrian crossing safety awareness in a signalized intersection pedestrian crossing scene; the detection model of pedestrian crossing safety consciousness of the signalless intersection is used for detecting pedestrian crossing safety consciousness in a pedestrian crossing scene of the signalless intersection; the detection model of the road section pedestrian crossing safety consciousness is used for detecting the pedestrian crossing safety consciousness in the road section pedestrian crossing scene;

The detection model of pedestrian crossing safety consciousness at the signalized intersection is as follows:

；

in the method, in the process of the invention, Scoring the safety awareness of pedestrians when crossing the street at the signalized intersection; Are all weight coefficients;

the detection model of pedestrian crossing safety awareness of no signalized intersection:

；

in the method, in the process of the invention, Scoring the safety awareness of pedestrians when crossing the street at the signalless intersection; Are all weight coefficients;

Detection model of road section pedestrian crossing safety consciousness:

；

in the method, in the process of the invention, : Grading the safety consciousness of pedestrians when crossing a road section; Are all weight coefficients;

step S5: grading the safety consciousness of the pedestrians according to the grading of the pedestrian crossing safety consciousness;

step S6: and controlling the vehicles with the nearby residual collision time with the pedestrians smaller than a preset value according to the safety awareness level of the pedestrians.

2. The vehicle driving control method considering pedestrian crossing safety awareness according to claim 1, wherein in step S2, the behavior parameters reflecting pedestrian crossing safety awareness can be extracted for a plurality of times or repeatedly used after first extraction according to the detection interval in the pedestrian crossing process; wherein the pedestrian selects the time point of crossing the street in the signal periodClose to the shelter from the pedestrian and plan for viewing time before crossing the streetIs extracted in the first detection period, and the first extraction result is used for each detection.

3. The vehicle driving control method considering pedestrian crossing safety awareness according to claim 2, wherein in step S2, the pedestrian' S perception-reaction time to the conflicting vehicleThe method comprises the steps of measuring the time for a pedestrian to sense and react to a vehicle with collision, wherein the parameter cannot be extracted when no vehicle with collision exists, and the value is 0 until the sensing-reacting time of the pedestrian to the collision vehicle is extracted for the first time, and if a new extraction result exists in the later detection, the new extraction result is used, otherwise, the average value of the previous extraction result is used; the remaining collision time of the pedestrian with the collision vehicle is used for reflecting whether the position or the action of the pedestrian is safe, and when no collision vehicle exists with the pedestrian, the remaining collision time of the pedestrian with the collision vehicle takes a value of +. Until the residual collision time of the pedestrian and the collision vehicle is extracted for the first time, if a new extraction result exists in the later detection, the new extraction result is used, otherwise, the average value of the previous extraction result is used.

4. A vehicle driving control method in consideration of pedestrian crossing safety awareness as set forth in claim 3, wherein in step S2, the remaining collision time of the pedestrian with the conflicting vehicleThe extraction method of (2) comprises the following steps: inputting the acquired pedestrian and vehicle data to the pedestrian crossingModel of pedestrian crossingThe model comprises a straight-going vehicle and a pedestrianModel, cornering vehicle and pedestrianA model; wherein, the straight-going vehicles and pedestrians areThe model is as follows:

；

；

；

；

；

in the method, in the process of the invention, Is the position of the vehicle; Is the position of the pedestrian; the longitudinal distance between the vehicle and the pedestrian at the initial moment; The transverse distance between the vehicle and the pedestrian at the initial moment; a longitudinal distance offset from the point of conflict; distance from front bumper to offset conflict point; And The length and width of the vehicle respectively; And The step length and the shoulder width of the pedestrian are respectively; is the speed of the vehicle; Is the speed of the pedestrian; is the lateral speed of the pedestrian; Is the longitudinal speed of the pedestrian;

turning vehicles and pedestrians The model is as follows:

；

；

；

；

；

；

in the method, in the process of the invention, Is the abscissa of the offset conflict point; is the position of the vehicle; Is the position of the pedestrian; And The length and width of the vehicle respectively; And The step length and the shoulder width of the pedestrian are respectively; is the speed of the vehicle; Is the speed of the pedestrian; is the lateral speed of the pedestrian; Is the longitudinal speed of the pedestrian; at the initial moment, the front bumper of the turning vehicle runs to the central angle of the initial conflict point; The central angle from the front bumper of the turning vehicle to the offset conflict point is used for driving; the longitudinal distance from the pedestrian to the initial conflict point is the initial moment; a longitudinal distance offset from the point of conflict; the longitudinal distance from the pedestrian to the offset conflict point; Is the central angle Corresponding arc distance; Is the central angle Corresponding arc distance; Is the turning radius of the vehicle.

5. The method according to claim 4, wherein in step S4, in the detection model of pedestrian crossing safety awareness of the signalized intersectionRespectively is; In the detection model of pedestrian crossing safety awareness of no-signal intersectionRespectively is; In the detection model of road section pedestrian crossing safety consciousnessRespectively is。

6. The vehicle driving control method considering pedestrian crossing safety awareness according to claim 5, wherein in step S5, 0-59 is classified into C-class, 60-79 is classified into B-class, 80-100 is classified into a-class, and if the pedestrian safety awareness level is a, the driver is alerted by voice that "pedestrian crossing in front of," please notice avoidance "; if the safety awareness level of the pedestrians is B, prompting a driver that the pedestrians cross the street in front through voice, please pay attention to avoid, and reducing the opening of the accelerator to be original; If the safety awareness level of the pedestrians is C, prompting a driver to 'walk ahead with pedestrians, please pay attention to avoid' through voice, and reducing the opening of an accelerator to be originalMeanwhile, the electronic braking force provided by the electronic hydraulic braking system is improved by 1.5 times.

7. The vehicle driving control system for realizing the method of claim 6, wherein the system comprises a vehicle data acquisition module, a vehicle-mounted communication module, a pedestrian data acquisition module, a road side communication module, a central control module, a data processing module, a pedestrian safety awareness calculation module, an accelerator control module, a brake control module and a voice control module;

the vehicle data acquisition module is used for acquiring the position and running state information of the vehicle;

the vehicle-mounted communication module is used for uploading running state information and technical parameters of the vehicle to the central control module and receiving a control strategy issued by the central control module;

the pedestrian data acquisition module is used for acquiring the position and speed of pedestrians and real-time street crossing videos;

The road side communication module is used for uploading pedestrian data to the central control module;

the central control module is used for receiving the data uploaded by the vehicle-mounted communication module and the road side communication module, acquiring real-time traffic flow data from a traffic management department and issuing a control strategy to the vehicle according to the safety awareness level of pedestrians; the pedestrian safety awareness calculation system comprises a data processing module and a pedestrian safety awareness calculation module;

The data processing module is used for processing pedestrian and vehicle data, extracting a series of behavior parameters capable of reflecting pedestrian crossing safety consciousness, and comprises the ratio of the gazing duration of pedestrians on signal lamps to the crossing duration Average distance of pedestrian from crosswalk centerlineHead rotation frequency in pedestrian crossing roadPerception-response time of pedestrian to conflicting vehiclesRemaining time of collision of pedestrian with conflicting vehicleThe time point of pedestrian crossing street in signal periodFixation point retention time during pedestrian crossingPedestrian approaches the shelter and plans the time of observation before crossing the street；

The pedestrian safety awareness calculation module is used for calling a corresponding safety awareness detection model according to a pedestrian crossing scene, and calculating the safety awareness score and grade of pedestrians when crossing the street by using the behavior parameters and the traffic flow parameters of the pedestrians;

the detection model of pedestrian crossing safety consciousness at the signalized intersection is as follows:

；

；

；

；

；

；

；

；

in the method, in the process of the invention, The fixation time length of the signal lamp is compared with the street crossing time length for pedestrians; the ratio of the gazing time length of the ideal pedestrian to the signal lamp to the street crossing time length is provided; standard deviation of ideal ratio; the average distance between the pedestrian and the center line of the crosswalk; Is the maximum allowable distance; A pedestrian selects a time point of crossing a street in a signal period in a pedestrian crossing scene at a signal intersection; The starting time of the green light is the starting time of the green light; the green light time is the green light time; h is the head rotation frequency in the process of pedestrian crossing; is the ideal head rotation frequency; Is the traffic volume; is the traffic flow speed; Is the time interval of the head; the head rotation frequency is used for adjusting the ideal head rotation frequency according to traffic volume, traffic flow speed and headway; the method is characterized in that the method is an adjusting coefficient, and the influence degree of traffic volume, traffic flow speed and headway on the ideal head rotation frequency of pedestrians is adjusted; a perceived-reaction time for a pedestrian to a conflicting vehicle; for adjusting the coefficient, the speed of increasing and decreasing the fraction is adjusted; remaining collision time for pedestrians and straight vehicles; Remaining collision time for pedestrians and turning vehicles; Ideal remaining collision time for pedestrians and conflicting vehicles; Scoring the safety awareness of pedestrians when crossing the street at the signalized intersection; Are all weight coefficients;

the detection model of pedestrian crossing safety awareness of no signalized intersection:

；

；

in the method, in the process of the invention, The pedestrian is on the way of crossing the streetDwell time of the individual gaze points; the total number of fixation points on the way of pedestrian crossing; is an ideal minimum residence time; is the ideal maximum residence time; to indicate a function, when the condition in brackets is satisfied, the value is 1, otherwise 0; Scoring the safety awareness of pedestrians when crossing the street at the signalless intersection; Are all weight coefficients;

Detection model of road section pedestrian crossing safety consciousness:

；

；

；

in the method, in the process of the invention, The method comprises the steps of approaching a shelter for pedestrians and planning the observation time before crossing a street; an ideal observation time before the pedestrian approaches the shelter and plans to cross the street; to adjust the coefficient, adjusting the rate of fractional reduction; Scoring the safety consciousness of pedestrians when crossing the road section; Are all weight coefficients;

the accelerator control module is used for controlling the opening degree of the accelerator;

the brake control module is used for controlling electronic assistance provided by the electronic hydraulic brake system;

the voice control module is used for controlling a voice system of the vehicle.

8. The vehicle driving control system according to claim 7, wherein the vehicle data acquisition module is connected to the vehicle-mounted radar camera to acquire vehicle running state information, and the vehicle running state information and the technical parameters are uploaded to the central control module through the vehicle-mounted communication module and stored in the first memory;

the pedestrian data acquisition module is connected with the millimeter wave radar and the high-definition camera arranged on the road side, acquires the position and the speed of the pedestrian and real-time street crossing video, uploads the pedestrian data to the central control module through the road side communication module, and stores the data on the second memory;

the central control module acquires real-time traffic flow data from a traffic management department and stores the real-time traffic flow data on a third memory;

the data processing module is used for retrieving vehicle data and pedestrian data from the first memory and the second memory, and extracting a series of behavior parameters capable of reflecting pedestrian crossing safety awareness by adopting a data processing method;

the pedestrian safety awareness calculating module invokes a corresponding safety awareness detecting model according to the pedestrian crossing scene, calculates the safety awareness score and grade of the pedestrian by combining the traffic flow data which are extracted from the third memory and the behavior parameters which are extracted by the data processing module, and stores the detecting result into the fourth memory;

and the central control module invokes the safety awareness level information of the pedestrians from the fourth memory and issues a vehicle control strategy according to the safety awareness level of the pedestrians.