CN106251572A - Smart Home based on vibration detection monitoring method and system - Google Patents

Abstract

The present invention relates to Smart Home technical field, it discloses a kind of Smart Home based on vibration detection monitoring method, comprise the steps: the structural vibration information that the behavioral activity of data collecting module collected people produces, carry out the behavioral activity of intelligent monitoring family members by analyzing and processing this signal and make alarm to abnormal.Beneficial effects of the present invention: solving the privacy concern of photographic head Indoor Video, the life on user does not brings impact, allows user use more easily.

Description

Smart home monitoring method and system based on vibration detection

technical field

The invention belongs to the technical field of smart home, in particular to an indoor positioning and monitoring method of smart home.

Background technique

With the development of the Internet of Things and the construction of smart cities, smart homes are becoming more and more popular. People hope that they can also monitor the situation at home when they are not at home, such as whether the elderly and children are safe, whether there are thieves invading, whether the babysitter has improper behavior towards children, etc. The current monitoring method is generally camera technology. However, installing cameras at home involves privacy and even legal issues, and it is even more impossible to install cameras in bedrooms and bathrooms, which are high-risk places where accidents happen to the elderly and children who are alone at home. In addition to privacy issues, the video captured by the camera continuously requires a huge amount of storage space, which leads to the overwriting of the video archive, making it impossible to obtain the video source before the storage period. In addition, due to the difficulty in realizing video self-identification classification and alarm technology, it is necessary to manually spend a lot of time watching videos to find the information people need when an accident occurs at home.

In view of the above reasons, many scientific researchers in the field of Internet of Things technology now devote themselves to WIFI-based human behavior detection. Modeling WIFI signals or using machine learning methods can identify human beings walking, running, falling, sitting, For behaviors such as brushing teeth and cooking, the indoor positioning of people can be realized by using the energy-decreasing characteristics or transmission speed of WIFI, and the walking track of people can be obtained. The advantage of WIFI lies in its popularity, which is available in almost every household. However, WIFI signals have multipath and non-line-of-sight problems. When separated by walls, the signal strength will drop sharply, which greatly affects the accuracy of indoor positioning and the recognition accuracy of human behavior. In addition, WIFI-based technology requires a fixed environment. If you move WIFI devices or change the indoor environment (such as the placement of furniture), this will cause the training model to no longer be valid, and behavior recognition will no longer be possible. Secondly, using WIFI to conduct behavior recognition on people can only have high accuracy when there is only one person. When other people or pets are present, the WIFI signal will be affected, resulting in failure to perform behavior recognition. Therefore, the technology of using WIFI to monitor indoors has not been commercialized.

Contents of the invention

In order to overcome the deficiencies in the prior art referred to above, the present invention provides a smart home monitoring method based on vibration detection, which collects and analyzes structural vibration signals caused by human behavior, especially ground vibration caused by footsteps. The vibration signal solves the shortcomings of the above-mentioned existing monitoring technology, and can locate, track and monitor indoor people.

The present invention is achieved through the following technical solutions:

A vibration detection-based smart home monitoring method includes the following steps: a data acquisition module collects structural vibration information generated by people's behaviors, and intelligently monitors the behaviors of family members by analyzing and processing the signals and gives an alarm for abnormalities.

As a further improvement of the present invention: the analysis process is specifically:

S1 analyzes and processes ground footsteps and vibration signals, identifies people, and issues alarms for illegal entry events;

S2 analyzes and processes ground footsteps and vibration signals, and uses TDOA three-point positioning to obtain precise location information of people;

S3 analyzes and processes vibration signals, pre-models or extracts features for different types of vibrations, and identifies human behavior.

As a further improvement of the present invention: human behavior is monitored by identifying vibration signals generated by human activities, and the vibration signals are analyzed and processed to identify human behavior.

As a further improvement of the present invention: the structural vibration is the ground vibration signal generated by people walking.

As a further improvement of the present invention: the data acquisition module includes a geophone, a pre-amplifier, a filter, a post-amplifier, an analog-to-digital conversion module and a microcontroller, and the data acquisition module includes a geophone, a pre-amplifier , a filter, a post-amplifier, an analog-to-digital conversion module and a microcontroller are connected in sequence.

As a further improvement of the present invention: the geophone acquires the direction angle of the footstep vibration signal, uses different angles to determine different footstep vibration signals, and then determines the number of people and achieves simultaneous positioning of multiple people.

The present invention also provides a smart home monitoring system based on vibration detection, including a data acquisition module, a data analysis module and a danger alarm module, the data acquisition module sends the collected structural vibration information to the data analysis module, and the data analysis module analyzes The results are sent to the hazard alert module.

As a further improvement of the present invention: the data acquisition module includes a geophone, a pre-amplifier, a filter, a post-amplifier, an analog-to-digital conversion module and a microcontroller, and the data acquisition module includes a geophone, a pre-amplifier , filter, post-stage amplifier, analog-to-digital conversion module and microcontroller are sequentially connected; the geophone obtains the direction angle of the footstep vibration signal, utilizes different angles to determine different footstep vibration signals, and then determines the number of people and Multiple people can be positioned at the same time.

As a further improvement of the present invention: the data analysis module performs indoor positioning, behavior detection and identification processing.

Beneficial effects of the present invention:

1. It solves the privacy problem of indoor camera monitoring, does not affect the life of users, and makes it more convenient for users to use.

2. Compared with video camera, the collected data is converted into digital text information by ADC for storage, which requires a small storage space and a long storage period.

3. Compared with video cameras, the collected data has been processed and analyzed, and there is no need to manually spend a lot of time watching videos. This invention can automatically detect accidents and send out alarms.

4. Vibration signals are transmitted through solids of the same material. Compared with electromagnetic wave signals such as WIFI, solids of the same material (such as floors) will not reflect or be blocked by walls. There is no non-line-of-sight and multipath effects. With a wide range and strong signal stability, more accurate original signals can be obtained, which is of great help to improve the accuracy of indoor positioning and behavior recognition in subsequent data analysis.

Description of drawings

Fig. 1 is a schematic structural diagram of a smart home monitoring system based on vibration detection in the present invention.

detailed description

The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

A vibration detection-based smart home monitoring method includes the following steps: a data acquisition module collects structural vibration information generated by people's behaviors, and intelligently monitors the behaviors of family members by analyzing and processing the signals and gives an alarm for abnormalities.

The analysis process is specifically:

S1 analyzes and processes ground footsteps and vibration signals, identifies people, and issues alarms for illegal entry events;

S2 analyzes and processes ground footsteps and vibration signals, and uses TDOA three-point positioning to obtain precise location information of people;

S3 analyzes and processes vibration signals, pre-models or extracts features for different types of vibrations, and identifies human behavior.

Monitor human behavior by identifying vibration signals generated by human activities, and analyze and process vibration signals to identify human behavior.

Structural vibration is the ground vibration signal generated by people walking.

The data acquisition module includes a geophone, a preamplifier, a filter, a postamplifier, an analog-to-digital conversion module and a microcontroller, and the data acquisition module includes a geophone, a preamplifier, a filter, a postamplifier , the analog-to-digital conversion module and the microcontroller are connected in sequence.

The geophone obtains the direction angle of the footstep vibration signal, and uses different angles to determine different footstep vibration signals, and then determines the number of people and enables multiple people to locate at the same time

The signal mainly collected by the present invention is the footstep vibration signal generated by the resident walking indoors. To upgrade occupant behavior recognition capabilities, seismometers can also be installed elsewhere to pick up structural vibrations due to human behavior.

Due to the difference in height and weight between people, body posture and walking habits, speed, etc., through the time-frequency analysis of footstep vibration signals, fast Fourier transform and wavelet transform technologies are used to extract features, and SVM classifiers are used. Machine learning technology can achieve the function of identifying people.

The present invention also provides a smart home monitoring system based on vibration detection, including a data acquisition module, a data analysis module and a danger alarm module, the data acquisition module sends the collected structural vibration information to the data analysis module, and the data analysis module analyzes The results are sent to the hazard alert module.

The data acquisition module includes a geophone, a preamplifier, a filter, a postamplifier, an analog-to-digital conversion module and a microcontroller, and the data acquisition module includes a geophone, a preamplifier, a filter, a postamplifier , the analog-to-digital conversion module and the microcontroller are connected in sequence; the geophone obtains the direction angle of the footstep vibration signal, and uses different angles to determine different footstep vibration signals, thereby determining the number of people and simultaneously positioning multiple people.

The data analysis module performs indoor positioning, behavior detection and identification processing.

When the user leaves home and needs to turn on the anti-theft mode, he can choose to call the police when someone enters the home and is active at home.

You can also set a more advanced anti-theft function, set a legal footstep vibration signal, and only call the police when a stranger enters.

Using the TDOA positioning algorithm, after more than three seismometers obtain footstep vibration signals, the occupant's location information can be calculated.

Through a variety of methods, it is possible to detect the number of people in the home and the indoor positioning of multiple people at the same time.

By pre-modeling different vibration types, or through machine learning, it is possible to classify different vibration types and identify human behavior, such as an old man falling.

When a fall-type vibration signal is detected, an alarm can be sent to the user's mobile phone, so that family members who are away from home can help the person who has fallen at home in time.

Combined with the classification technology of people's location information, more behavior categories can be identified, such as sleeping, cooking, going to the toilet, etc.

When family members sleep for more than a period of time in a row, it can also be set as an unexpected alarm event and sent to the remote user.

Since most of people's indoor activities will contact and vibrate with objects in the home, such as the ground, stove, etc., by training or modeling more vibration signals, the functions can be continuously upgraded to recognize more and more human behaviors .

The invention also provides a smart home monitoring system based on vibration detection, which is composed of three modules. Data acquisition module, data processing module and danger alarm module. The data acquisition module can choose a geophone (but not limited to a geophone). A geophone is a highly sensitive instrument for detecting vibrations, which is more sensitive than a vibration sensor.

After decades of technological development, geophones are relatively mature and low in cost, and can be deployed in large areas.

The data acquisition module is composed of a geophone, a preamplifier, a filter, a postamplifier, an analog-to-digital conversion module, and a microcontroller.

Since the unprocessed signal directly collected by the geophone is too weak, an amplifier is needed to improve the signal-to-noise ratio. For noise, we need filters to denoise. Use the analog-to-digital converter ADC to convert the analog signal into a digital signal, and finally send it to the microcontroller to prepare for data processing and analysis.

The data processing module is mainly divided into three functions: indoor positioning, behavior detection and identification.

For the original signal, the notch filter is used to remove the current noise, and the Wiener filter and other techniques are used to improve the signal-to-noise ratio. It uses fast Fourier transform and wavelet transform technology to process, and then cuts the required signal with segmentation algorithm.

Indoor Positioning:

Through the TDOA algorithm, the location information of a person can be obtained by using three-point positioning, and the specific location of the occupant can be known. Continuous detection can obtain the walking trajectory of family members.

The microcontroller sets a higher sampling frequency. Due to the signal propagation advantage of solid vibration, there is no problem of multipath and non-line-of-sight, and the positioning accuracy of several centimeters can be achieved.

When there is more than one person in the family, if the footstep vibration signals of different people overlap and overlap, because the characteristics of the footstep vibration of multiple people and the duration of the footstep vibration signal of a single person are different, through training, use machine learning According to the classification method, the number of people in the family can be known.

The signal is collected by a three-component geophone group, and the direction angle of the footstep vibration signal can be obtained. Using different angles, different footstep vibration signals can be determined, so as to determine the number of people and even locate multiple people at the same time.

In addition to the above two methods, there are a variety of signal separation methods, which can achieve the effect of simultaneous positioning of multiple people.

In addition, when people in the family are scattered and active, the detectors with the closest distance will obtain signals with higher energy values. Through the selection algorithm, only three detectors with high energy values are selected respectively, and the location information of people can be calculated independently. Multi-person positioning can achieve the effect of single-person positioning.

Behavior detection:

First, train some basic vibration types, such as the vibration signal of a fall, and obtain a training model. When the collected vibration signal matches the fall sample, it is considered a fall event.

Here is a classification method for reference, but there are many classification methods: use Mel cepstrum coefficient (MFCC) as a feature, and establish a corresponding probability model for each individual in the feature set of each vibration signal through a Gaussian mixture model , the distribution of the individual features of each vibration signal in the feature space is abstracted as the random result of the probability model, and the EM algorithm is used to estimate the GMM parameters so that the logarithmic likelihood function has a maximum value, and the logarithmic likelihood function can be obtained through testing. Likelihood value, and the similarity of its simulation can be measured by the range of logarithmic likelihood value. Various vibration signals can be classified by the range of log-likelihood values.

Because many behaviors at home are limited to specific places, such as cooking in the kitchen and going to the toilet in the bathroom. Therefore, combining location information and walking trajectory, more behavioral activities can be identified.

In order to upgrade the behavior detection function, seismometers can be installed not only on the ground, but also in more places where humans may come into contact, and more human behaviors can be identified by training vibration signals.

Identification:

Each person has a unique walking pattern due to many factors, such as the individual's physical characteristics, the position of the center of gravity when walking, the way the feet contact the ground, etc. In the time domain, frequency domain, or a combination of time and frequency, the unique vibration characteristics of footsteps produced by each person can be found. Using these features, identification can be achieved.

Because the duration of each person's footsteps is different from the interval of footsteps, it is also a feasible solution to use this point as a feature, and then use the KNN classifier for identification.

Danger alarm module has occupant accidental alarm and burglar intrusion alarm.

When the occupants have some unexpected dangerous behaviors, such as falling, sleeping for a long time, etc., an alarm will be sent.

When an illegal identity appears at home within the set time, an alarm is issued.

The alarm can push the alarm message to the user through SMS or mobile app.

The above content is a further elaboration of the present invention in combination with specific implementation methods, and it should not be assumed that the specific implementation of the present invention is limited to the above description. For those skilled in the art, without departing from the concept of the present invention, some simple deductions or substitutions can be made, which should be deemed to be within the scope of protection determined by the claims submitted in the present invention.

Claims (9)

1. A smart home monitoring method based on vibration detection, it is characterized in that: comprise the steps: the structural vibration information that data acquisition module gathers people's behavioral activity produces, come intelligent monitoring family member's behavioral activity by analyzing and processing this signal and to Exception alerts. 2. The smart home monitoring method based on vibration detection according to claim 1, wherein the analysis process is specifically: S1 analyzes and processes ground footsteps and vibration signals, identifies people, and issues alarms for illegal entry events; S2 analyzes and processes ground footsteps and vibration signals, and uses TDOA three-point positioning to obtain precise location information of people; S3 analyzes and processes vibration signals, pre-models or extracts features for different types of vibrations, and identifies human behavior. 3. The smart home monitoring method based on vibration detection according to claim 2, characterized in that, the human behavior is monitored by identifying vibration signals generated by human activities, and the vibration signals are analyzed and processed to identify human behavior. 4. the smart home monitoring method based on vibration detection according to claim 1, is characterized in that: structural vibration is the ground vibration signal that people walk and produce. 5. the smart home monitoring method based on vibration detection according to claim 1, is characterized in that: described data acquisition module comprises geophone, preamplifier, filter, postamplifier, analog-to-digital conversion module and microcontroller The data acquisition module includes a geophone, a preamplifier, a filter, a postamplifier, an analog-to-digital conversion module and a microcontroller connected in sequence. 6. the smart home monitoring method based on vibration detection according to claim 5, is characterized in that: geophone obtains the direction angle of footstep vibration signal, utilizes different angles, determines different footstep vibration signals, and then determines the number of people And do multiple people positioning at the same time. 7. A smart home monitoring system based on vibration detection, characterized in that: it includes a data acquisition module, a data analysis module and a danger alarm module, the data acquisition module sends the collected structural vibration information to the data analysis module, and the data analysis module sends The analysis results are sent to the danger alert module. 8. The smart home monitoring system based on vibration detection according to claim 7, characterized in that: the data acquisition module includes a geophone, a preamplifier, a filter, a postamplifier, an analog-to-digital conversion module and a microcontroller The data acquisition module includes a geophone, a pre-amplifier, a filter, a post-amplifier, an analog-to-digital conversion module, and a microcontroller connected in sequence; the geophone obtains the direction angle of the footstep vibration signal, and utilizes different The angle of the footsteps can be determined to determine the vibration signals of different footsteps, and then determine the number of people and achieve simultaneous positioning of multiple people. 9. The smart home monitoring system based on vibration detection according to claim 7, characterized in that: the data analysis module performs indoor positioning, behavior detection and identification processing.