Real-time road safety assessment method based on running vehicle vibration sensing, medium and equipment

Abstract

The invention provides a real-time road safety assessment method based on driving vehicle vibration sensing, a medium and equipment, and the method comprises the steps: firstly dividing a road needing to be monitored into regions according to a specified size, and dividing the road needing to be monitored into sections according to the region attribution and the road type; establishing a neural network regression model for predicting a vibration difference value according to the road type, the temperature data, the weather condition and the road surface obstacle; and finally, comparing a vibration sensing value generated in the driving process of the automobile with a historical standard vibration sensing value of the automobile after a series of standardization processing, thereby helping to judge whether abnormal conditions such as cracks, recesses and lower structure changes occur on the road or not. According to the invention, manual detection cost can be effectively reduced, risks can be found in time, countermeasures can be taken, and road accidents can be avoided.

Description

Technical Field

[0001] The present invention belongs to the field of road traffic safety, and in particular relates to a real-time road safety assessment method, medium and equipment based on vibration sensing of a moving vehicle. Background Art

[0002] With the continuous expansion of modern transportation systems and the large-scale use of vehicles, road safety and maintenance have become an important part of public management. Road damage and potential collapse risks not only threaten the personal safety of drivers and passengers, but may also cause serious traffic delays and economic losses. Traditional road detection methods usually rely on regular manual inspections and expensive professional equipment. These methods are not only time-consuming and labor-intensive, but also often fail to monitor road conditions in real time, making it difficult to detect potential safety hazards in a timely manner, so as to take early warning measures and avoid accidents. Summary of the invention

[0003] In view of the deficiencies in the prior art, the present invention provides a real-time road safety assessment method, medium and device based on vibration sensing of a moving vehicle.

[0004] To achieve the above object, the present invention adopts the following technical solutions:

[0005] In a first aspect, the present invention provides a real-time road safety assessment method based on vibration sensing of a moving vehicle, characterized by comprising:

[0006] S1: Divide the roads to be monitored into areas of specified size. The divided areas are called points. Each point records the latitude and longitude range it contains. Divide the roads to be monitored into sections according to the area affiliation and road type. Each section records the point information it contains and the road type it belongs to.

[0007] S2: On normal interval roads, the latitude and longitude, vibration sensor values ​​and temperature data of multiple vehicles traveling at speeds within a specified range for multiple days are collected in real time, and photos of the road conditions in front of the vehicles are taken in real time. Weather conditions and road obstacles are identified based on the photos of the road conditions in front of the vehicles;

[0008] S3: Each longitude and latitude and its corresponding road type, temperature data, weather conditions and road obstacles are used as a set of input data and input into the neural network regression model for training. The label of the neural network regression model is the absolute value of the difference between the average value of the historical vibration sensor value and the measured vibration sensor value of the point corresponding to the longitude and latitude, which is recorded as the vibration difference value ΔF;

[0009] S4: Collect the time, longitude and latitude GPS, vibration sensor value F, temperature data T, road condition photos ahead and vehicle speed uploaded in real time by vehicles on the road to be evaluated, exclude the uploaded data whose speed is not within the specified range, and after the car has completed a section, match the longitude and latitude uploaded by the car in the section with the point position to determine the data corresponding to the point position;

[0010] S5: extract the data corresponding to each point and input it into the trained neural network regression model to predict the vibration difference value ΔF, use the vibration difference value ΔF to supplement the vibration sensor value F, and obtain the standard vibration sensor value F′;

[0011] S6: By comparing the standard vibration sensor value F′ with the N historical vibration sensor values ​​of the point, it is determined whether the vibration sensor value of the point is abnormal, and the abnormality is processed.

[0012] Optionally, in step S1, a unique id is used to identify the point, and the point set is represented by Pid=(pid 1 ,pid 2 , ..., pid k ), k is the number of points; a unique id is used to identify the interval, and the interval set is represented by Aid = (aid 1 ,aid 2 ,...,aid l ), l is the number of intervals; record the road type Rid = (rid 1 ,rid 2 , ..., rid n ), n is the number of road types; record the relationship between Pid and Aid.

[0013] Optionally, in step S2, the car is equipped with a vibration sensor, a temperature sensor and a front-end camera of the car, which are respectively used to collect vibration sensing values, temperature data and photos of road conditions in front of the car.

[0014] Optionally, in step S3, the weather conditions and road obstacles are one-hot encoded before being input into the neural network regression model; the corresponding road type is found according to each longitude and latitude, and each road type, temperature data, weather code, and road obstacle code are used as a set of input data and input into the neural network regression model for training.

[0015] Optionally, in step S4, after the car completes a section, all uploaded data of the car in the section are extracted, and the time, longitude and latitude, vibration sensor value, temperature data, vehicle speed, weather code, and road obstacle code uploaded every second are taken as a group of data sets, and valid data sets that meet the speed requirements are screened, and the longitude and latitude are extracted from the valid data sets. The corresponding road type and point are found according to the longitude and latitude, and this group of data sets belongs to the data set within the point; if there are multiple groups of data sets within the point, the average value of each type of data is taken to represent the data set corresponding to the point.

[0016] Optionally, in step S5, the vibration sensor value F=(F 1 , F 2 , ..., F m ) and the corresponding point Pid′=(pid 1 ,pid 2 , ..., pid m ), m is the number of data; at the same time, the standard vibration sensor values ​​corresponding to the N Pid′ closest to the current day are extracted from the past historical data of the car. i∈[1, N]; input the road type, temperature data, weather code and road obstacle code corresponding to Pid′ into the trained neural network regression model, and predict a set of vibration difference values ​​ΔF=(ΔF 1 , ΔF 2 , ..., ΔF m ); This group of ΔF is used as the supplementary value of the vibration sensor value F of the day, and the standard vibration sensor value F′=(F 1 +ΔF 1 , F 2 +ΔF 2 , ..., F m +ΔF m )=(F′ 1 , F′ 2 , ..., F′ m ).

[0017] Optionally, in step S6, the isolation forest algorithm is used to compare F′ and F′ i Conduct data analysis to determine whether there is any abnormal vibration sensor value at any point on that day; if there is an abnormal point, mark the point; if multiple vehicles have abnormalities at a certain point, issue an alarm forecast for that point.

[0018] Optionally, after step S6, the method further includes using the valid data set acquired on the day as a training set and inputting it into the neural network regression model for training again.

[0019] In a second aspect, the present invention provides a computer-readable storage medium storing a computer program, characterized in that the computer program enables a computer to execute the real-time road safety assessment method based on vibration sensing of a moving vehicle as described in the first aspect.

[0020] In a third aspect, the present invention provides an electronic device, characterized in that it includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the real-time road safety assessment method based on vibration sensing of a moving vehicle as described in the first aspect is implemented.

[0021] The beneficial effects of the present invention are as follows: the present invention utilizes the vibration sensor values ​​monitored by the vehicle during driving, and standardizes the vibration sensor values ​​to eliminate the influence of the vehicle's own factors and external environmental factors on the vehicle's vibration sensor values; and compares the real-time standard vibration sensor values ​​of the vehicle with the previous standard vibration sensor values ​​to evaluate the structural integrity and potential collapse risk of the road. The present invention has the advantages of strong real-time performance, low labor cost, and wide coverage, and can effectively predict the occurrence of road accidents, facilitate timely response, and improve the safety of road use. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 The data processing flow chart of a real-time road safety assessment method based on vibration sensing of a moving vehicle.

[0023] Figure 2 The figure is a flow chart of the overall method of a real-time road safety assessment method based on vibration sensing of a moving vehicle.

[0024] Figure 3 The flowchart of the training prediction model of the present invention is explained by taking a single car as an example. DETAILED DESCRIPTION

[0025] The technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application.

[0026] In one embodiment, if Figure 1 and Figure 2 As shown, the present invention proposes a real-time road safety assessment method based on vibration sensing of a moving vehicle, which specifically includes the following steps.

[0027] Step 1: Install vibration sensors, temperature sensors and front-end cameras on the car.

[0028] Step 2: Divide the road to be monitored into multiple areas with a unit of 10 meters × 10 meters. These areas are called points, and use unique ids to identify these points. The point set can be expressed as Pid = (pid 1,pid 2 , ..., pid k ), and at the same time, record the latitude and longitude range of each point. In addition, the road to be monitored is divided into intervals according to factors such as regional ownership and road type (asphalt pavement, cement concrete pavement, etc.), and these intervals are identified with a unique id. The interval set can be expressed as Aid = (aid 1 ,aid 2 ,...,aid l ), and at the same time, record the road type Rid = (rid 1 ,rid 2 , ..., rid n ). Finally, record the relationship between Pid and Aid, such as pid 1 ,pid 2 ...,pid u Belong to aid 1 etc., where u∈(1, k). The above data are saved in a record table.

[0029] Step 3: During the early training process, a large number of cars are allowed to drive at a speed within the specified range on normal interval roads (no cracks, no obvious bumps, etc.) for many consecutive days. During the driving process, real-time latitude and longitude GPS, vibration sensor value F, temperature data T, and real-time photos of the road conditions in front of the car taken by the front camera of the car are uploaded at a speed of one second. At the same time, the back-end will use artificial intelligence algorithms to identify weather and obstacles to process the uploaded photos and identify the weather conditions and road obstacles on the current road. The weather can be divided into eight conditions: sunny, light rain, moderate rain, heavy rain, light snow, heavy snow, strong winds, and sandstorms. These conditions are based on the degree to which the weather will affect the vibration of the vehicle. The road obstacles are considered based on the types that cars usually run over when driving, so they are divided into foam boards, speed bumps, pebbles, small branches, mud and sand, and plastic bags.

[0030] Step 4: One-hot encode the weather conditions and road obstacle conditions respectively. The encoding details are shown in Table 1 and Table 2. Enter the corresponding encoding value according to the actual identified weather type and road obstacle type. Then find the corresponding road type rid in the record table according to each GPS, and use the weather code and road obstacle code corresponding to each rid, T, and photo recognition result as a set of input data, and input them into the neural network regression model for training. The relevant vibration data of the GPS corresponding point is processed as the model label, which is Each time a set of data is input into the regression model, all the collected data is input into the model separately. After training, the model can predict the vibration difference value ΔF of the same car at the same point caused by external environmental factors such as road type, temperature, weather and road obstacles. The training process is as follows Figure 3 shown.

[0031] Table 1 Weather codes

[0032] sunny 1 0 0 0 0 0 0 0 light rain 0 1 0 0 0 0 0 0 moderate rain 0 0 1 0 0 0 0 0 heavy rain 0 0 0 1 0 0 0 0 Light Snow 0 0 0 0 1 0 0 0 heavy snow 0 0 0 0 0 1 0 0 Strong wind 0 0 0 0 0 0 1 0 Sandstorm 0 0 0 0 0 0 0 1

[0033] Table 2 Road obstacle codes

[0034] Foam board 1 0 0 0 0 0 Speed ​​bumps 0 1 0 0 0 0 Little stone 0 0 1 0 0 0 Small branches 0 0 0 1 0 0 Sediment 0 0 0 0 1 0 plastic bag 0 0 0 0 0 1

[0035] Step 5: In actual application, every day, the car driving on the road to be evaluated continuously uploads its real-time time S, longitude and latitude GPS, vibration sensor value F, temperature data T, real-time road condition photos ahead and car speed v at a speed of one second. The speed V is used to detect whether the car's driving speed is within the specified range. Only when the car's speed is within the specified range, its data will be used as data for subsequent analysis of road abnormalities. This method can effectively eliminate the phenomenon that the car's driving speed is large due to factors such as speeding and traffic jams, which leads to large differences in vibration sensor values. When a car runs through a section, extract all the data uploaded by the car in this section, and take the S, GPS, F, T, V, weather code, and road obstacle code uploaded every second as a set of data sets. If V is not within the specified range, the set of data is invalid. At the same time, extract the GPS whose v meets the requirements, find the corresponding rid and point pid in the record table according to the GPS information, and the set of data belongs to a set of data sets corresponding to the point pid; if there are multiple sets of data with very close time in a certain point, the value of the data is represented by the average value under the same type of data, for example Where v is the number of sets of data.

[0036] Step 6: Extract the vibration sensor value F = (F 1 , F 2 , ..., F m ) and the corresponding point Pid′=(pid 1 ,pid 2 , ..., pid m ), where m represents the number of valid data sets for the car in this interval. At the same time, the standard vibration sensor values ​​corresponding to the 10 Pid′ closest to the current day are extracted from the car’s past historical data. i∈[1,10]. If there are less than N data points in the history, the actual amount of data will be used; if there is no historical data for some points before the day, the point will not be included in the subsequent analysis, but the processed standard vibration sensor value of the point on the day will be stored in the historical data of the point. The rid, T, weather code and road obstacle code corresponding to the photo recognition result of the day corresponding to Pid′ are input into the neural network regression model respectively, and finally a set of vibration difference values ​​ΔF=(ΔF 1 , ΔF 2 , ..., ΔF m ), and at the same time, this group of ΔF is used as the supplementary value of the vibration sensor value F of the day, and the two are added together to obtain the standard vibration sensor value F′=(F 1 +ΔF 1 , F 2 +ΔF 2 , ..., F m +ΔF m )=(F′ 1 , F′ 2 , ..., F′ m ). The predicted ΔF is added to the F value of the point to which the data input into the regression model belongs.

[0037] Step 7: Finally, use the isolation forest algorithm to find F′=(F′ 1 , F′ 2 , ..., F′ m )and Perform data analysis to determine whether the vibration sensor value of any point on that day is abnormal, that is, too high or too low. If there is an abnormal point, mark the point. The core idea of ​​the isolation forest algorithm is that in normal data distribution, data points tend to cluster together, while abnormal points are relatively isolated. Therefore, by randomly splitting the data space, abnormal points will be isolated in a separate subspace earlier. In the algorithm analysis process, the current day and historical data of the same point are analyzed as a group of values, that is, There are m sets of values ​​in total. It is worth noting that the value of the day is always placed at the end of a set of data. The process of analyzing anomalies is as follows:

[0038] 1. The sample set input by the algorithm each time is Since the number of sample sets is small, the number of constructed trees should not be set too high, and the reference range is [5, 20] for training.

[0039] 2. Set the F′ of the day j Input it into the trained isolation forest model and calculate its anomaly score. The formula is:

[0040]

[0041] Among them, E(h(F′ j )) is the data point F′ j The average path length in all trees, t is the sample size of the training set, and c(t) is the average path length of the tree. The range of the anomaly score is [0, 1]. The closer the score is to 1, the easier it is to isolate the sample and the greater the possibility of an outlier.

[0042] 3. If the anomaly score is greater than 0.5, the value is judged to be an outlier and the point corresponding to the value is recorded.

[0043] Step 8: All data uploaded by cars are processed by car and interval according to the methods of steps 5, 6 and 7. The standard vibration sensor values ​​finally obtained by different cars in different intervals are only recorded in their own databases. It should be noted that each car has a dedicated database, which stores the standard vibration sensor values ​​F′ of the car at different points in different intervals at different times. Each F′ corresponds to an interval aid, point pid and time S.

[0044] Step 9: If multiple vehicles have abnormal values ​​at a certain point, an alarm will be issued for that point.

[0045] Step 10: Pass the newly obtained data set into the neural network regression model again to continuously train and adjust the accuracy of the model.

[0046] In another embodiment, the present invention provides a computer-readable storage medium storing a computer program, wherein the computer program enables a computer to execute the real-time road safety assessment method based on vibration sensing of a moving vehicle as described in the first embodiment.

[0047] In another embodiment, the present invention proposes an electronic device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the real-time road safety assessment method based on vibration sensing of a moving vehicle as described in Example 1 is implemented.

[0048] In the embodiments disclosed in the present application, the computer storage medium can be a tangible medium that can contain or store a program for use by or in conjunction with an instruction execution system, device or equipment. The computer storage medium can include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the above. More specific examples of computer storage media can include electrical connections based on one or more lines, portable computer disks, hard disks, random access memories (RAM), read-only memories (ROM), erasable programmable read-only memories (EPROM or flash memory), optical fibers, portable compact disk read-only memories (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the above.

[0049] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed in this application can be implemented in electronic hardware or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0050] The above are only preferred embodiments of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions under the concept of the present invention belong to the protection scope of the present invention. It should be pointed out that for ordinary technicians in this technical field, some improvements and modifications without departing from the principle of the present invention should be regarded as the protection scope of the present invention.

Claims

1. A real-time road safety assessment method based on vibration sensing of a moving vehicle, characterized in that: include: S1: Divide the roads to be monitored into areas of specified size. The divided areas are called points. Each point records the latitude and longitude range it contains. Divide the roads to be monitored into sections according to the area affiliation and road type. Each section records the point information it contains and the road type it belongs to. S2: On normal interval roads, the latitude and longitude, vibration sensor values ​​and temperature data of multiple vehicles traveling at speeds within a specified range for multiple days are collected in real time, and photos of the road conditions in front of the vehicles are taken in real time. Weather conditions and road obstacles are identified based on the photos of the road conditions in front of the vehicles; S3: Each longitude and latitude and its corresponding road type, temperature data, weather conditions and road obstacles are used as a set of input data and input into the neural network regression model for training. The label of the neural network regression model is the absolute value of the difference between the average value of the historical vibration sensor value and the measured vibration sensor value of the point corresponding to the longitude and latitude, which is recorded as the vibration difference value ΔF; S4: Collect the time, longitude and latitude GPS, vibration sensor value F, temperature data T, road condition photos ahead and vehicle speed uploaded in real time by vehicles on the road to be evaluated, exclude the uploaded data whose speed is not within the specified range, and after the car has completed a section, match the longitude and latitude uploaded by the car in the section with the point position to determine the data corresponding to the point position; S5: extract the data corresponding to each point and input it into the trained neural network regression model to predict the vibration difference value ΔF, use the vibration difference value ΔF to supplement the vibration sensor value F, and obtain the standard vibration sensor value F′; S6: By comparing the standard vibration sensor value F′ with the N historical vibration sensor values ​​of the point, it is determined whether the vibration sensor value of the point is abnormal, and the abnormality is processed.

2. A real-time road safety assessment method based on vibration sensing of a moving vehicle as claimed in claim 1, characterized in that: In step S1, a unique id is used to identify the point, and the point set is represented by Pid=(pid1, pid2, ..., pid k ), k is the number of points; a unique id is used to identify the interval, and the interval set is represented by Aid = (aid1, aid2, ..., aid ι ), l is the number of intervals; record the road type of each interval Rid = (rid1, rid2, ..., rid n ), n is the number of road types; record the relationship between Pid and Aid.

3. A real-time road safety assessment method based on vibration sensing of a moving vehicle as claimed in claim 1, characterized in that: In step S2, the car is equipped with a vibration sensor, a temperature sensor and a front-end camera for collecting vibration sensing values, temperature data and photos of the road conditions in front of the car, respectively.

4. A real-time road safety assessment method based on vibration sensing of a moving vehicle as claimed in claim 1, characterized in that: In step S3, the weather conditions and road obstacles are one-hot encoded before being input into the neural network regression model; the corresponding road type is found according to each longitude and latitude, and each road type, temperature data, weather code, and road obstacle code are used as a set of input data and input into the neural network regression model for training.

5. A real-time road safety assessment method based on vibration sensing of a moving vehicle as claimed in claim 1, characterized in that: In step S4, after the car has completed a section, all the data uploaded by the car in the section are extracted, and the time, longitude and latitude, vibration sensor value, temperature data, vehicle speed, weather code, and road obstacle code uploaded every second are taken as a set of data sets, and the valid data sets that meet the speed requirements are screened, and the longitude and latitude are extracted from the valid data sets. The corresponding road type and point are found according to the longitude and latitude, and this set of data sets belongs to the data set within the point; if there are multiple sets of data sets within the point, the average value of each type of data is taken to represent the data set corresponding to the point.

6. A real-time road safety assessment method based on vibration sensing of a moving vehicle as claimed in claim 5, characterized in that: In step S5, the vibration sensor value F=(F1, F2, ..., F m ) and the corresponding point Pid′=(pid1,pid2,...,pid m ), m is the number of data; at the same time, the standard vibration sensor values ​​corresponding to the N Pid′ closest to the current day are extracted from the past historical data of the car. The road type, temperature data, weather code and road obstacle code corresponding to the day of Pid′ are input into the trained neural network regression model, and a set of vibration difference values ​​ΔF=(ΔF1, ΔF2, ..., ΔF m ); This group of ΔF is used as the supplementary value of the vibration sensor value F of the day, and the standard vibration sensor value F′ is obtained by adding the values ​​to eliminate the influence of external environmental factors. m +ΔF m )=(F′1, F′2,...,F′ m ).

7. A real-time road safety assessment method based on vibration sensing of a moving vehicle as claimed in claim 6, characterized in that: In step S6, the isolation forest algorithm is used to compare F′ and F′ i Conduct data analysis to determine whether there is any abnormal vibration sensor value at any point on that day; if there is an abnormal point, mark the point; if multiple vehicles have abnormalities at a certain point, issue an alarm forecast for that point.

8. A real-time road safety assessment method based on vibration sensing of a moving vehicle as claimed in claim 5, characterized in that: After step S6, the process also includes taking the valid data set acquired on the day as a training set and inputting it into the neural network regression model for training again.

9. A computer-readable storage medium storing a computer program, characterized in that: The computer program enables the computer to execute the real-time road safety assessment method based on vibration sensing of a moving vehicle as described in any one of claims 1-8.

10. An electronic device, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the real-time road safety assessment method based on vibration sensing of a moving vehicle as described in any one of claims 1 to 8 is implemented.

Images