CN116080679A - An unmanned driving method for trackless rubber-tyred vehicles facing underground roadways - Google Patents

Abstract

The invention belongs to the technical field of unmanned driving, and discloses an unmanned driving method for trackless rubber-tyred vehicles facing underground roadways. The method includes: accurate in the early stage, task determination, and roadway perception; receiving sensed and detected information through the state decision-making layer, Determine the driving state according to the detected information; the control decision-making layer selects the corresponding control decision according to the determined driving state; the tracking control layer selects the corresponding tracking control method according to the selected control decision. This method can adapt to the special environmental conditions in the mine, and the perception results can be used for different state decision-making judgments and control execution switching. While ensuring the control accuracy, it can flexibly handle various dynamic scenarios such as meeting vehicles, bypassing obstacles, and stopping obstacles, reducing the number of manual interventions and improving Unmanned driving operation efficiency, reduce system cost.

Description

An unmanned driving method for trackless rubber-tyred vehicles facing underground roadways

technical field

The invention belongs to the technical field of unmanned driving, and in particular relates to an unmanned driving method for trackless rubber-tyred vehicles facing underground roadways.

Background technique

As an important part of the country's green mine construction, like open pit mines, the development of unmanned driving technology in underground mines is also particularly critical. Compared with open-pit mines, underground mines have limited communication, high positioning costs, and high requirements for decision-making control, making it difficult to realize unmanned driving. At present, for unmanned driving in underground roadways, there are control methods that rely on the combination of vehicle sensors and road sensors, which can realize horizontal and vertical control in static scenes, but the installation cost of road sensors is high; there are also trackless rubber-tyred vehicles based on high-precision maps. The driving planning control method, due to the accurate map information and real-time positioning information, can achieve more accurate control, but the amount of data for the mapping of narrow and long roadways is large and the cycle is long. The existing research results can initially meet the needs of unmanned driving of underground trackless rubber-tyred vehicles, but the cost, preparation period, and processing capacity in dynamic environments still need to be improved. Therefore, the unmanned driving of underground trackless rubber-tyred vehicles has not yet been widely used .

Contents of the invention

In view of the above-mentioned problems in the prior art, the present invention proposes an unmanned driving method for trackless rubber-tyred vehicles facing underground roadways.

The technical scheme that the present invention adopts is specifically as follows:

An unmanned driving method for trackless rubber-tyred vehicles facing underground roadways, comprising the following steps:

S1, pre-preparation: set up a signboard at the fork in the roadway, the signboard has characteristic information and its corresponding curvature information that can be recognized by sensors to distinguish different forks;

Curvature signboards are placed at large curvature curves, and the large curvature curves are curves whose curvature exceeds the range of curvature that can be identified by the sensor;

Place a mission end sign at the end of the path;

Sensors installed on the vehicle end for sensing information in the roadway;

S2, determine the task: determine the overall driving route according to the actual task requirements in the mine; determine the characteristic information on the identification plate that needs to be recognized at each fork that the driving route passes through according to the determined overall driving route;

S3, roadway perception: perceive the roadway boundary information in real time, and perform secondary fitting, and send the boundary abnormal state to the state decision-making layer for state decision-making;

Real-time perception and output of obstacle information to the state decision-making layer for state decision-making;

Real-time perception and output of detected curvature signage information to the decision-making layer for state decision-making;

The inertial navigation outputs the slope information to the tracking control layer for tracking control;

S4, state decision-making: make driving state decision-making according to the obtained perception information;

In the state decision-making process, different driving states are judged through the output results of roadway perception, and the judged driving state is output to the control decision-making layer for control decision-making;

S5, control decision-making, the control decision-making layer processes the boundary information differently according to the different driving states of the state decision-making layer, and sends the final boundary information to the tracking control layer;

S6, tracking control, select the corresponding control mode according to the received slope information and boundary information; wherein, the slope information is used for longitudinal control, and the vertical direction outputs the accelerator/brake pedal opening based on the fuzzy calibration table, and Add ramp supplementary amount; the boundary information is used for lateral control, including bilateral control, unilateral control, central control and parking control;

S7, the task is over, the infrared camera recognizes the task end sign, decelerates to stop and clears the corner, and ends; and/or, the parking control exceeds a given time, sends a message that it cannot continue to pass to the control center, requests manual takeover, and is currently driving automatically end of task

Otherwise, according to the real-time perception information of the sensor, keep the state switching control in steps S3-S6.

Further, in the state decision-making process, the driving state includes:

In the normal driving state, when it is detected that the boundaries on both sides of the roadway are normal, no obstacles are detected;

In the obstacle avoidance driving state, when it is detected that the boundaries on both sides of the roadway are normal, obstacles are detected, and the obstacles can be bypassed;

Abnormal driving state, when the boundary of one side of the roadway is missing or the boundary of both sides is missing, no obstacle is detected;

In the curve driving state, when the roadway curvature identification information is detected, no obstacles are detected;

Take over the parking state, the vehicle is faulty, the roadway boundary is normal and obstacles cannot be bypassed, the roadway boundary is missing and obstacles are detected, the roadway curvature identification information is detected and obstacles are detected;

Wherein, the curve driving state has priority over the abnormal driving state and the normal driving state.

最后将当前坐标转换至后轴中心，利用跟踪算法计算前轮转角控制量；左侧边界拟合曲线为y＝a₁x²+b₁x+c₁，右侧边界拟合曲线为y＝a₂x²+b₂x+c₂。Further, for roadway information perception, when the boundaries on both sides of the detected roadway are normal and no obstacles are detected, and the state is judged to be in a normal driving state, the control decision is normal tracking, and the tracking control is bilateral control: the preview is determined by the preview distance Point P(x _p ,y _p ), bring x _p into the left and right boundary fitting curves, get y ₁ , y ₂ and calculate the target point

Finally, convert the current coordinates to the center of the rear axle, and use the tracking algorithm to calculate the control amount of the front wheel angle; the left boundary fitting curve is y=a ₁ x ² +b ₁ x+c ₁ , and the right boundary fitting curve is y= a ₂ x ² +b ₂ x+c ₂ .

Further, for roadway information perception, when the boundaries on both sides of the detected roadway are normal and obstacles are detected, when the state is judged to be the obstacle avoidance driving state, the control decision is abnormal tracking, and the tracking control is unilateral control: determine the predicted distance through the preview distance. Aim at point P(x _p ,y _p ), bring x _p into the boundary fitting curve on the side far away from the obstacle, get y ₁ and calculate the target point Q(x _p ,y ₁ -d), d is the boundary Offset the safety distance, and finally convert the current coordinates to the center of the rear axle, and use the tracking algorithm to calculate the control amount of the front wheel angle; if the obstacle is in the middle of the roadway, choose one side of the boundary to fit the curve.

Further, roadway information perception, when one side of the roadway boundary is normal, the other side of the roadway boundary is abnormal, no obstacles are detected, and the state is determined to be an abnormal driving state, the control decision is abnormal tracking, and the tracking control is unilateral control: by The preview distance determines the preview point P(x _p ,y _p ), brings x _p into the boundary fitting curve on one side in the normal state, obtains y ₁ and calculates the target point Q(x _p ,y ₁ -d) , d is the safety distance of the boundary offset, and finally the current coordinates are converted to the center of the rear axle, and the control amount of the front wheel angle is calculated by using the pure tracking algorithm.

Further, in roadway information perception, when abnormalities are detected on both sides of the roadway and no obstacles are detected, the state is judged as an abnormal driving state, the control decision is abnormal tracking, and tracking control is the central control: normal boundary parameter fitting at the previous moment Based on the central curve y=a ₀ x ² +b ₀ x+c ₀ , the preview point P(x _p ,y _p ) is determined by the preview distance, and the target point Q is calculated by bringing x _p into the center fitting curve (x ₀ ,y ₀ ), and finally convert the current coordinates to the center of the rear axle, and use the pure tracking algorithm to calculate the control amount of the front wheel angle.

Further, roadway information perception, the roadway perception module detects the curvature identification information in the roadway information signboard, and when the state judgment changes to driving on a curve, the control decision is curve tracking. At this time, the boundaries on both sides are fitted to the center curve, according to The best bend-in point of curvature fitting and the identified curvature are spliced with the center curve and the circular trajectory; the tracking control selects the center control.

Further, roadway information perception, when the vehicle fails, obstacles cannot be detoured, the roadway boundary is missing and obstacles are detected, the roadway curvature identification information is detected and obstacles are detected, and the state is determined to take over the parking state, the control decision is to stop Wait, the tracking control enters into the parking control;

Parking control: After the longitudinal accelerator opening gradually decays to zero, the brake opening gradually increases. When the vehicle speed is zero, parking is enabled, and the pedal opening and steering wheel angle are initialized.

Compared with the prior art, the beneficial effects of the present invention are:

1. The present invention provides an unmanned driving method for trackless rubber-tyred vehicles facing underground roadways, which can adapt to the special environmental conditions in the mine. The sensing results can be used for different state decision-making judgments and control execution switching, ensuring control accuracy and flexible handling A variety of dynamic scenarios such as meeting cars, bypassing obstacles, and stopping obstacles can reduce the number of manual interventions, improve the efficiency of unmanned driving operations, and reduce system costs.

In addition, the present invention does not rely on road surface sensors and roadway sensors for the underground roadway unmanned driving method, but relies on vehicle-mounted multi-sensors for environmental perception, and the control unit switches control modes according to environmental information to complete unmanned driving.

2. The present invention places digital signboards through the curvature information measured in advance, identifies signboards with different digital features according to task requirements during actual operation, and performs boundary fitting predictions to realize stable unmanned driving of vehicles at fork roads and improve system complexity. Environmental adaptability.

3. The present invention aims at abnormal situations such as unilateral boundary loss, bilateral boundary loss, and large curvature that cannot be sensed, and solves the problem of unmanned driving control stability in perception abnormal scenes through methods such as unilateral boundary optimization, midline fitting, and curvature fitting. , Improve the operating efficiency and safety of the unmanned driving system.

Description of drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the accompanying drawings that need to be used in the embodiments, and the features and advantages of the present invention will be more clearly understood by referring to the accompanying drawings , the accompanying drawings are schematic and should not be construed as limiting the present invention in any way. For those skilled in the art, other drawings can be obtained according to these drawings without creative work. in:

Fig. 1 is a control logic schematic diagram of the present invention;

Fig. 2 is the schematic diagram of decision-making control in the process of vehicle entering curve and branch road of the present invention (wherein part a is to perceive signboard information to enter the curve at the roadway curve, adopting central control; part b is to identify at the roadway branch position Signage information, select the curve path according to the mission path, and use the central control when entering the curve);

Fig. 3 is a schematic diagram of tracking path optimization in the state of missing boundary during vehicle driving of the present invention (the state where the bilateral boundary of part a is normal and there is an obstacle ahead; part b is the state of one-sided boundary and there is no obstacle ahead; c is the state where the bilateral boundary is abnormal and there is no obstacle in front);

Figure 4 is a schematic diagram of different boundary control principles;

Explanation of reference signs: P preview point, Q target point, d boundary offset safety distance, O point is the steering center of the vehicle.

Detailed ways

In order to understand the above-mentioned purpose, features and advantages of the present invention more clearly, the present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that, in the case of no conflict, the embodiments of the present invention and the features in the embodiments can be combined with each other.

In the following description, many specific details are set forth in order to fully understand the present invention. However, the present invention can also be implemented in other ways different from those described here. Therefore, the protection scope of the present invention is not limited by the specific details disclosed below. EXAMPLE LIMITATIONS.

Example 1

This embodiment discloses an unmanned driving method for trackless rubber-tyred vehicles facing underground roadways, which includes roadway perception, state decision-making, control decision-making, and tracking control. Roadway perception is based on vehicle-mounted lidar and infrared cameras, and outputs driving boundary information, road curvature information, and obstacle information. State decision-making determines different driving states based on perceptual information. The control decision outputs different tracking information according to different driving states. The tracking control calculates the control quantity based on the tracking information and finally controls the vehicle to drive along the roadway path.

With reference to accompanying drawing 1 shown, a kind of unmanned driving method of trackless rubber-tyred vehicle facing underground roadway, comprises the following steps:

S1, pre-preparation: set up a signboard at the roadway fork, the signboard has infrared features and its corresponding curvature information that can be recognized by sensors to distinguish different forks;

Curvature signs are placed at large curvature curves; the basis for judging "large curvature" in this embodiment is to determine whether the radar in the perception module can normally detect the boundary during driving, as a person skilled in the art should What is known is that when the curvature of the curve exceeds the detection range of the configured radar, that is, the radar cannot normally sense it, that is, the curvature is too large, which is called a large curvature; in this embodiment, the large curvature is curvature>0.03. That is, "large curvature" is a curvature beyond the detection range of the radar.

Place a mission end sign at the end of the path;

The vehicle end is equipped with sensors for sensing information in the roadway.

S2, determine the task: according to the actual task requirements in the mine, determine the overall driving route; according to the determined overall driving route, determine the characteristic information on the signboard that needs to be identified at each fork that the overall driving route passes through, in this embodiment , the feature information is a digital feature. That is, according to the overall driving route, the nth fork of the route needs to identify the digital feature (infrared feature) on the signboard, that is, the nth fork will drive towards the fork in the preset direction after recognizing the specific infrared feature.

S3, roadway perception: LiDAR perceives the roadway boundary information in real time, and performs secondary fitting, and sends the boundary abnormal state to the state decision-making layer for state decision-making;

Laser radar and infrared camera fusion output obstacle information to the state decision-making layer for state decision-making, infrared camera real-time output detected curvature signage information to the state decision-making layer for state decision-making, inertial navigation outputs ramp information to the tracking control layer for control decision-making ;

S4, state decision: make a decision on the driving state, the driving state includes normal driving, abnormal driving, obstacle avoidance driving, curve driving and takeover parking; normal driving means that the boundary is normal and no obstacles are detected; obstacle avoidance driving means that the boundary is normal However, obstacles are detected and can be detoured; when driving on a curve, the infrared camera detects the curvature signboard information and no obstacles are detected; taking over parking means that the vehicle fails, the roadway boundary is normal and obstacles cannot be detoured, and the roadway boundary is missing and Obstacles are detected, roadway curvature identification information is detected, and any obstacle is detected;

In the state decision-making process, different driving states are judged through the roadway perception output results, and the judged driving states are output to the control decision-making layer;

S5, control decision, control decision includes normal tracking, abnormal tracking, curve tracking, parking waiting; normal tracking corresponds to normal driving state; abnormal tracking corresponds to obstacle avoidance driving and abnormal driving state; curve tracking corresponds to curve driving state, curve During the tracking process, curvature fitting is performed based on the curvature information and the current boundary information; parking waiting corresponds to the upper layer taking over the parking state, and the control output is initialized during the parking waiting process;

The control decision-making layer processes the boundary information differently according to the different decision-making states of the state decision-making layer, and sends the final boundary information to the tracking control layer.

S6, tracking control, select the corresponding control mode according to the received slope information and boundary information; wherein, the slope information is used for longitudinal control, and the vertical direction outputs the accelerator/brake pedal opening based on the fuzzy calibration table, and Add ramp supplementary amount; the boundary information is used for lateral control, and lateral control is divided into bilateral control, unilateral control, central control and parking control according to different states; bilateral control calculates the control amount according to the issued bilateral boundary, The unilateral control tracks the processed unilateral information, the central control tracks the issued centerline, and the parking control is to decelerate to stop and clear the corner after receiving the control decision to stop and wait for the order.

S7, the task is over, the infrared camera recognizes the task end sign, decelerates to stop and resets the corner to zero; and/or, the parking control exceeds a given time, sends a message that it cannot continue to pass to the control center, and requests manual takeover, the current automatic driving task Finish;

Otherwise, maintain the state switching control in steps S3-S6 according to the real-time sensing information of the sensor.

specifically:

Roadway perception performs boundary detection, obstacle detection, and curvature recognition of bends and turns, and outputs the results to the state decision-making part. State decision-making includes normal driving, abnormal driving, obstacle avoidance driving, curve driving, taking over the parking state, and jumps according to the perception information; control decision-making includes normal tracking, abnormal tracking, curve tracking, parking waiting state, and corresponding Jump; Tracking control adopts corresponding control methods according to the upper-level decision, including two-sided control, one-sided control, central control, and parking control. The specific implementation process is as follows:

Curvature fitting, referring to Figure 2, in this embodiment, for each branch in the mine and large curvature curves that cannot be sensed by the laser radar, the manual driving is collected and converted into the front wheel angle through the steering wheel angle, and the average is calculated according to the curvature and angle formula. Curvature, determine the best bend-in point (that is, the starting point S of the fitting path in Figure 2).

Placement of signboards. In this embodiment, signboards with curvatures of different infrared reflection characteristics are placed at the fork in accordance with left, middle and right. The first part of the number on the signboard represents the direction, "1" is left, "2" is middle, " 3" is right, the second part of the number ranges from 0 to 999, which is 1000 times the corresponding curvature (that is, the resolution is 0.001), such as "1080", that is, the curvature of the left-turn curve is 0.08 (the specific installation position can be seen in Figure 2). Curvature signboards are also placed at large curvature curves, the first part of the number is fixed as "0", and the second part of the number is the same as the definition of the fork; a signboard with the number "9999" is placed at the end of the mission.

The task is determined. According to the actual operation requirements in the mine, the appropriate driving route is selected, and the parameters of the digital signature signs that the roadway sensing part needs to recognize at the fork are configured. This parameter consists of a series of numbers, and each number corresponds to the first part of the signboard that needs to be identified, such as "2-1-3-1-2", which means going straight, turning left, and turning right at the fork in the whole process , go straight and turn right.

Longitudinal control, the longitudinal output of the accelerator/brake pedal opening is based on the fuzzy calibration table, and the amount of slope compensation is added. Due to the use of a mature algorithm, no explanation is given here.

The following controls correspond to horizontal controls in different states:

Boundary detection control, the task starts to execute, and the roadway perception first performs boundary detection, based on the lidar point cloud data to extract the boundary within 10m ahead, and fit it into a quadratic curve in the radar coordinate system (radar is the origin, along the axis of the vehicle is the x-axis , perpendicular to the x-axis, the right is the y-axis), and consider the basic characteristics of the boundary to judge the abnormality of the fitting curve. If it is abnormal, change the value of the curve flag to 0, and the normal value is 1. The format is as follows:

Table 1 Boundary fitting curves on both sides

direction flag bit Curve fitting left side 1 <![CDATA[y=a₁x²+b₁x+c₁]]> Right 1 <![CDATA[y=a₂x²+b₂x+c₂]]>

最后将当前坐标转换至后轴中心，利用纯跟踪算法计算前轮转角控制量。When the boundaries on both sides are normal and no obstacles are detected, the state decision is normal driving; the control decision is normal tracking; the tracking control is bilateral control: as shown in Figure 4 a, the preview point is determined by the preview distance P(x _p ,y _p ), bring x _p into the left and right boundary fitting curves, get y ₁ , y ₂ and calculate the target point

Finally, the current coordinates are converted to the center of the rear axle, and the control amount of the front wheel angle is calculated by using the pure tracking algorithm.

When the boundary on one side of the detected roadway boundary is normal, the boundary on the other side is abnormal, and no obstacles are detected, the state decision is abnormal driving; the control decision is abnormal tracking, as shown in b in Figure 3; the tracking control is unilateral control: As shown in b in Figure 4, the preview point P(x _p ,y _p ) is determined by the preview distance, and x _p is brought into the normal boundary fitting curve to obtain y ₁ and calculate the target point Q(x _p ,y ₁ -d), where d is the safe distance of boundary offset, and finally convert the current coordinates to the center of the rear axle, and use the pure tracking algorithm to calculate the control amount of the front wheel angle.

When the boundaries on both sides of the detected roadway are abnormal and no obstacles are detected, the state decision is abnormal driving; the control decision is abnormal tracking, as shown in c in Figure 3, and the tracking control is central control: as shown in c in Figure 4, Based on the center curve y=a ₀ x ² +b ₀ x+c ₀ fitted by the normal boundary parameters at the previous moment, the preview point P(x _p ,y _p ) is determined by the preview distance, and x _p is brought into the center The target point Q(x ₀ , y ₀ ) is calculated by fitting the curve, and finally the current coordinates are converted to the center of the rear axle, and the control amount of the front wheel angle is calculated by using the pure tracking algorithm.

Obstacle sensing control, in this embodiment, because the underground mainly relies on the lidar sensing boundary positioning, the obstacle avoidance driving state is only triggered when both sides of the boundary exist and obstacles can be bypassed. During the obstacle avoidance driving process, the roadway sensing part will output the fusion sensing results of lidar and infrared camera, output obstacle information, and realize obstacle bypassing and meeting vehicles.

After detecting an obstacle, judge whether it can pass through the obstacle and the boundary position. In the actual unmanned driving environment, especially on the unmanned operating route, in order to ensure transportation efficiency, parking caused by obstacles will be avoided as much as possible Control happens. Therefore, take obstacle circumvention as an example (if it is judged that it can pass, the obstacle circumvention will be performed):

As shown in Figure 3a, when an obstacle is detected on the right side, the boundaries on both sides of the roadway are normal at this time, the state decision is obstacle avoidance driving, and the control decision is abnormal tracking, as shown in Figure 3b; the tracking control is One-sided control, as shown in b in Figure 4, the subsequent control principle adopts the one-sided control in the above-mentioned boundary detection control.

Curve turnout recognition control, in the curvature fitting work, has been calibrated for the large curvature boundary that cannot be recognized by the lidar (the corresponding identification information board is preset in the roadway), and the curvature identification board at the curve is recognized during the driving of the vehicle Or the curvature signboard at the branch road, the state decision is changed to curve driving, and the control decision is changed to curve tracking, as shown in a in Figure 2. At this time, the boundaries on both sides are fitted to the central curve y=a ₀ x ² +b ₀ x+c ₀ , and according to the best bending point of curvature fitting and the identified curvature, splicing the center curve and the circular trajectory; the tracking control is the center control, and the control principle adopts the above-mentioned center control.

When the vehicle drives to the fork, the roadway perception accurately identifies the corresponding entrance signboard and curvature according to the initially input task information, the state decision becomes curve driving, and the control decision becomes curve tracking, as shown in b in Figure 2. The center curve splicing and center control in the process are the same as those in the curve driving process.

Other state control, when the vehicle breaks down, obstacles cannot be detoured, roadway boundaries are missing and obstacles are detected, roadway curvature identification information is detected and obstacles are detected, and the state is determined to take over the parking state, the control decision is to stop and wait , the tracking control directly enters the parking control. After the longitudinal accelerator opening gradually decays to zero, the brake opening gradually increases. When the vehicle speed is zero, parking is enabled, and the pedal opening and steering wheel angle are initialized.

Furthermore, the foregoing are merely illustrative of some embodiments, and changes, modifications, additions and/or variations may be made without departing from the scope and spirit of the disclosed embodiments, which are illustrative and not restrictive. Furthermore, the illustrated embodiments relate to what are presently considered to be the most practical and preferred embodiments, it is to be understood that the embodiments should not be limited to the disclosed embodiments, but rather are intended to cover the spirit and scope of the embodiments included. Different modifications and equivalent settings within . In addition, various implementations described above can be used together with other implementations, for example, aspects of one implementation can be combined with aspects of another implementation to implement yet another implementation. Additionally, individual features or components of any given assembly may constitute additional embodiments.

The foregoing description of the embodiments has been provided for purposes of illustration and description, and is not intended to be exhaustive or to limit the present disclosure. Elements or features of a particular embodiment are generally not limited to that particular embodiment, but, where applicable, are interchangeable and can be used in a selected embodiment, even if not specifically shown or described, and also Can be changed in many ways. Such variations are not to be regarded as a departure from the disclosure, and all such variations are included within the scope of the disclosure.

Therefore, it should be understood that the drawings and description are provided herein by way of example to facilitate the understanding of the present invention and should not be construed as limiting the scope thereof.

Claims (8)

1. The unmanned method of the trackless rubber-tyred vehicle facing the underground roadway is characterized by comprising the following steps of:

s1, preparation in the early stage: a signboard is arranged at a turnout in a roadway, and the signboard is provided with characteristic information which can be identified by a sensor and is used for distinguishing different turnouts and corresponding curvature information thereof;

placing a curvature signboard at a large curvature curve, wherein the large curvature curve is a curve with curvature exceeding the curvature range which can be identified by the sensor;

placing a task ending signboard at a path end point;

the vehicle end is provided with a sensor for sensing information in the roadway;

s2, determining a task: determining an overall driving path according to the underground actual task demand; according to the determined overall driving path, determining characteristic information on a signboard which needs to be identified at each turnout on the driving path;

s3, roadway sensing: sensing roadway boundary information in real time, performing secondary fitting, and sending the boundary abnormal state to a state decision layer for state decision;

sensing and outputting obstacle information to a state decision layer in real time to make a state decision;

sensing and outputting the detected curvature signboard information to a state decision layer in real time to perform state decision;

the inertial navigation outputs ramp information to a tracking control layer for tracking control;

s4, state decision: carrying out running state decision according to the obtained perception information;

in the state decision process, different driving states are judged according to the roadway-perceived output result, and the judged driving states are output to a control decision layer for control decision;

s5, controlling decision, wherein the control decision layer carries out different processing on the boundary information according to different running states of the state decision layer, and sends the final boundary information to the tracking control layer;

s6, tracking control, namely selecting a corresponding control mode according to the received ramp information and boundary information; the ramp information is used for longitudinal control, the throttle/brake pedal opening degree is longitudinally output based on a fuzzy calibration table, and the ramp supplement quantity is added; the boundary information is used for transverse control, including double-side control, single-side control, central control and parking control;

s7, finishing the task, identifying a task finishing signboard by the infrared camera, decelerating and stopping, and clearing a corner; and/or the parking control exceeds a given time, sending the information that the vehicle cannot continue to pass to a control center, requesting manual taking over, and ending the current automatic driving task;

otherwise, according to the real-time sensing information of the sensor, the state switching control in the S3-S6 steps is maintained.

2. The unmanned method of trackless rubber-tyred vehicle for underground roadway of claim 1, wherein,

in the state decision process, the driving state comprises:

in a normal running state, when the two side boundaries of the roadway are detected to be normal, no obstacle is detected;

in the obstacle avoidance driving state, when the normal boundary of the two sides of the roadway is detected, an obstacle is detected, and the obstacle can be bypassed;

in an abnormal driving state, when the boundary deletion at one side or both sides of the roadway is detected, no obstacle is detected;

a curve driving state, when roadway curvature identification information is detected, no obstacle is detected;

taking over a parking state, wherein a vehicle breaks down, a roadway boundary is normal and an obstacle cannot bypass, the roadway boundary is absent and the obstacle is detected, roadway curvature identification information is detected and the obstacle is detected;

wherein the curve running state is prioritized over the abnormal running state and the normal running state.

3. A trackless rubber-tyred vehicle unmanned method for a down-hole roadway according to claim 1 or 2, wherein,

roadway information perception, when the boundaries on two sides of a detected roadway are normal and no obstacle is detected, and when the state is judged to be a normal running state, a control decision is normal tracking, and tracking control is double-side control: determination of a pretightening point P (x) by pretightening distance _p ,y _p ) Will x _p Bringing the left and right boundary fitting curves to obtain y ₁ 、y ₂ And calculate the target point

Finally, converting the current coordinates to the center of the rear axle, and calculating the control quantity of the front wheel steering angle by using a tracking algorithm; left boundary fit curve y=a ₁ x ² +b ₁ x+c ₁ Right boundary fit curve is y=a ₂ x ² +b ₂ x+c ₂ 。

4. The unmanned method of the trackless rubber-tyred vehicle facing the underground roadway according to claim 1 or 2, wherein roadway information is perceived, when the boundaries on two sides of the detected roadway are normal and an obstacle is detected, and the state is judged to be an obstacle avoidance driving state, a control decision is abnormal tracking, and tracking control is single-side control: determination of a pretightening point P (x) by pretightening distance _p ,y _p ) Will x _p Carrying out boundary fitting curve on one side far away from the obstacle to obtain y ₁ And calculate the target point Q (x _p ,y ₁ D), d is a boundary offset safety distance, and finally the current coordinate is converted to the center of the rear axle, and the control quantity of the front wheel steering angle is calculated by using a tracking algorithm; and if the obstacle is positioned in the middle of the roadway, optionally fitting a curve on one side boundary.

5. The unmanned method of a trackless rubber-tyred vehicle facing an underground roadway according to claim 1 or 2, wherein the roadway information is perceived, when one side boundary of the roadway is normal, the other side boundary of the roadway is abnormal, no obstacle is detected, and the state is judged to be an abnormal driving state, the control decision is abnormal tracking, and the tracking control is single-side control: determination of a pretightening point P (x) by pretightening distance _p ,y _p ) Will x _p Carrying out a side boundary fitting curve in a normal state to obtain y ₁ And calculate the target point Q (x _p ,y ₁ D), d is a boundary offset safety distance, and finally the current coordinate is converted to the center of the rear axle, and the pure tracking algorithm is utilized to calculate the steering angle control quantity of the front wheel.

6. A trackless rubber-tyred vehicle unmanned method for a down-hole roadway according to claim 1 or 2, wherein,

roadway information perception, namely judging an abnormal driving state when detecting that the boundaries on two sides of a roadway are abnormal and no obstacle is detected, wherein a control decision is abnormal tracking, and tracking control is central control: the center curve y=a of the normal boundary parameter fitting at the above time ₀ x ² +b ₀ x+c ₀ On the basis of this, a pretightening point P (x _p ,y _p ) Will x _p The target point Q (x) is calculated by taking in the center fitting curve ₀ ,y ₀ ) And finally, converting the current coordinate to the center of the rear axle, and calculating the control quantity of the front wheel steering angle by using a pure tracking algorithm.

7. A trackless rubber-tyred vehicle unmanned method for a down-hole roadway according to claim 1 or 2, wherein,

the roadway information sensing module detects curvature identification information in the roadway information identification plate, when the state judgment is changed into curve driving, the control decision is curve tracking, at the moment, the boundaries on two sides are fitted into a central curve, and the central curve and the circular track are spliced according to the optimal bending point fitted by the curvature and the identified curvature; the tracking control selects the center control.

8. The unmanned method of trackless rubber-tyred vehicle for underground roadway according to claim 2, wherein,

roadway information sensing, namely when a vehicle fails, an obstacle cannot bypass, a roadway boundary is absent, the obstacle is detected, roadway curvature identification information is detected, the obstacle is detected, and the state is judged to be in a parking taking state, a control decision is parking waiting, and tracking control enters parking control;

and (3) parking control: after the longitudinal accelerator opening gradually decays to zero, the brake opening gradually increases, the parking is enabled when the vehicle speed is zero, and the pedal opening and the steering wheel angle are initialized.

Images