Risk Control Big Data Mining Method and System Based on Internet of Vehicles

Abstract

The present invention provides a risk control big data mining method and system based on the vehicle networking. First, real-time operation data streams of a target vehicle in the vehicle networking environment are collected, including vehicle state time series data, driving behavior event sets, external environment perception data, and in-vehicle communication interaction records. Then, combined operation feature vectors, abnormal behavior pattern identifiers, and dynamic environment risk factors are respectively extracted from different data, and a structured communication feature matrix is generated and transformed. Based on this, a spatio-temporal association risk analysis map is constructed. Through the spatio-temporal association risk analysis map, a dynamic driving risk assessment is carried out on the target vehicle to generate a risk level quantification index and a risk event warning label. Finally, a risk control decision instruction set is generated according to the above results and fed back to the in-vehicle control terminal and the vehicle networking cloud platform to achieve a comprehensive assessment and effective control of vehicle risks.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle networking, and in particular, to a risk control big data mining method and system based on vehicle networking. Background Art

[0002] In the context of the rapid development of the current intelligent transportation field, vehicle networking technology has gradually become a research and application hotspot, aiming to improve traffic efficiency and safety through information interaction between vehicles and the external environment, and provide more intelligent and convenient services for drivers.

[0003] Currently, in terms of vehicle risk management in a vehicle networking environment, most of the existing technologies stay at simple statistics and classification of some common driving behaviors, and it is difficult to deeply mine potential abnormal behavior patterns. These methods often do not consider the mutual correlation between different driving behaviors and the dynamic changes of behaviors in the time dimension, and cannot identify abnormal driving behaviors that may lead to safety risks in a timely and accurate manner. Summary of the Invention

[0004] In view of the problems mentioned above, in combination with the first aspect of the present invention, embodiments of the present invention provide a risk control big data mining method based on vehicle networking, and the method includes:

[0005] Collect real-time operation data streams generated by a target vehicle in a vehicle networking environment, where the real-time operation data streams include vehicle state time series data, driving behavior event sets, external environment perception data, and in-vehicle communication interaction records;

[0006] Extract combined operation feature vectors from the vehicle state time series data, extract abnormal behavior pattern identifiers from the driving behavior event sets, parse dynamic environment risk factors from the external environment perception data, and convert the in-vehicle communication interaction records into structured communication feature matrices;

[0007] Based on the combined operation feature vectors, the abnormal behavior pattern identifiers, the dynamic environment risk factors, and the structured communication feature matrices, construct a spatio-temporal correlation risk analysis map;

[0008] Dynamically evaluate the driving risks of the target vehicle through the spatio-temporal correlation risk analysis map, and generate risk level quantification indicators and risk event warning labels;

[0009] Generate a risk control decision instruction set according to the risk level quantification indicators and the risk event warning labels, and feedback the risk control decision instruction set to the in-vehicle control terminal and the vehicle networking cloud platform of the target vehicle.

[0010] In another aspect, an embodiment of the present invention further provides a risk control big data mining system based on the vehicle Internet of Things, including a processor and a machine-readable storage medium. The machine-readable storage medium is connected to the processor. The machine-readable storage medium is used to store programs, instructions or codes, and the processor is used to execute the programs, instructions or codes in the machine-readable storage medium to implement the above method.

[0011] Based on the above aspects, after collecting the multi-dimensional real-time operation data stream of the target vehicle in the vehicle Internet of Things environment, the embodiments of the present application respectively extract key features from different types of data, including combined operation feature vectors, abnormal behavior pattern identifiers, dynamic environment risk factors, and structured communication feature matrices, and then construct a spatio-temporal correlation risk analysis map therefrom, so as to deeply fuse and correlate analyze multi-source heterogeneous data, accurately capture the complex spatio-temporal risk relationships during the vehicle operation process, and show the evolution law of the vehicle operation risk in the time and space dimensions. Moreover, using the constructed spatio-temporal correlation risk analysis map to dynamically evaluate the driving risk of the target vehicle can generate risk level quantification indicators and risk event warning labels in real time, and can adjust the risk assessment results in a timely manner according to the real-time operation situation of the vehicle, providing timely and accurate risk information for drivers and relevant management platforms, and effectively coping with the continuously changing risk situations during the vehicle driving process. Finally, generate a risk control decision instruction set based on the risk level quantification indicators and risk event warning labels, and feedback it to the vehicle-mounted control terminal and the vehicle Internet of Things cloud platform. This enables the vehicle Internet of Things system to quickly make targeted decisions according to the risk assessment results, realize active intervention and effective control of vehicle risks, ensure the safe driving of vehicles, and improve the safety and reliability of the entire vehicle Internet of Things ecosystem. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 It is a schematic execution flow diagram of a risk control big data mining method based on the vehicle Internet of Things provided by an embodiment of the present invention.

[0013] Figure 2 It is a schematic diagram of exemplary hardware and software components of a risk control big data mining system based on the vehicle Internet of Things provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0014] The present invention will be specifically described below with reference to the accompanying drawings of the specification. Figure 1 It is a schematic flow diagram of a risk control big data mining method based on the vehicle Internet of Things provided by an embodiment of the present invention. The risk control big data mining method based on the vehicle Internet of Things will be introduced in detail below.

[0015] Step S110, collect the real-time operation data stream generated by the target vehicle in the vehicle networking environment, where the real-time operation data stream includes vehicle status time-series data, driving behavior event sets, external environment perception data, and in-vehicle communication interaction records.

[0016] Specifically, taking an electric vehicle driving on an urban road as an example of the target vehicle, in the vehicle networking environment, the electronic control unit (ECU) of the vehicle continuously records various data to form a real-time operation data stream. In terms of vehicle status time-series data, the engine speed sequence records the speed changes of the electric vehicle's motor. For example, when the vehicle starts, the motor speed gradually increases, remains stable during uniform driving, and rapidly decreases during braking. The braking pressure waveform reflects the pressure changes during each braking operation. The battery temperature curve changes with factors such as vehicle driving time, charge-discharge cycles, and environmental temperature. The tire wear increment gradually accumulates with the increase in driving mileage. In the driving behavior event set, the operation events of the driver are recorded, such as the driver performing a sudden acceleration operation when the vehicle in front suddenly decelerates, frequently making consecutive lane changes on a multi-lane road, speeding in certain sections, and not using lights as required at night. In terms of external environment perception data, the sensors equipped on the vehicle sense real-time meteorological condition parameters. For example, when it rains, the rain drop sensor detects the rainfall intensity and the visibility decreases; the camera identifies the heat map of road obstacle distribution. For example, there is a vehicle parked on the side of the road due to a breakdown on the road, which is shown as a high-risk obstacle aggregation area on the heat map; at the same time, it can also accurately obtain traffic sign semantic information, such as identifying that the speed limit sign ahead requires the vehicle speed not to exceed 60 kilometers per hour, and the movement trajectory sets of surrounding vehicles, such as the driving directions, speeds, and distances from the target vehicle of surrounding vehicles. The in-vehicle communication interaction records include the communication data between the vehicle and other nearby vehicles and roadside infrastructure (such as intelligent traffic lights, roadside units), such as the vehicle obtaining information such as the remaining time of the traffic light by communicating with the traffic light.

[0017] Step S120, extract the combined operation feature vectors from the vehicle status time-series data, extract the abnormal behavior pattern identifiers from the driving behavior event sets, analyze the dynamic environment risk factors from the external environment perception data, and convert the in-vehicle communication interaction records into a structured communication feature matrix.

[0018] Specifically, for the engine speed sequence in the vehicle state time-series data, when performing frequency-domain transformation, for example, the periodic fluctuation amplitude characteristics of the motor speed during the acceleration process of an electric vehicle can reflect the working stability of the motor. If the speed mutation detection marker shows a sudden speed drop during normal driving, it may indicate a malfunction in the motor or transmission system. When performing piecewise integration on the brake pressure waveform to generate a brake intensity distribution histogram, if it is found that the intensity of most braking operations is concentrated in the lower region, but there are occasional abnormal pressure fluctuation thresholds for high-intensity braking, this may indicate abnormalities in the driver's driving habits or the braking system. After trend decomposition of the battery temperature curve, the long-term temperature rise gradient characteristic shows that the battery temperature continues to rise and has a large slope during long-term driving, and the short-term temperature oscillation parameter shows that the battery temperature has rapid small-amplitude fluctuations under certain working conditions, which may affect the battery life and performance. After modeling the wear rate of the tire wear increment, the calculated tire life prediction index shows that the remaining tire life is short and the probability of abnormal tread wear is high, which may be due to the vehicle driving on roads with poor road conditions for a long time or bad driving habits. Thus, the above features can be normalized and spliced to form a combined operation feature vector.

[0019] When extracting abnormal behavior pattern identifiers from the driving behavior event set, filter out the number of times the hard acceleration event is triggered. For example, if the hard acceleration event is triggered 5 times within an hour, after comparing with the preset acceleration threshold, a higher hard acceleration risk coefficient is generated. The statistical result of the time interval of consecutive lane-changing events shows that the time interval of the driver's consecutive lane-changing on a section of road is very short, and a higher lane-changing behavior density score is determined. The proportion of the overspeed duration is calculated to be 30% of the total driving time during a section of the journey, and a large overspeed violation cumulative amount is generated in combination with the road speed limit data. The frequency of illegal use of lights at night is relatively high, and a light violation pattern identifier is generated by matching with the light specification database. These data are weighted and fused to generate an abnormal behavior pattern identifier.

[0020] Parse dynamic environmental risk factors from external environment perception data, identify that the real-time meteorological condition parameter is rainy weather, calculate that the visibility attenuation coefficient is relatively high, and the road adhesion risk index also increases due to rain. The heat map of road obstacle distribution shows that there is a broken-down vehicle on the right side of the road ahead. The obstacle collision probability distribution map generated based on its density gradient indicates that there is a relatively high collision probability when the target vehicle approaches the broken-down vehicle. Conduct compliance verification on the semantic information of traffic signs. If it is found that the vehicle is about to speed, generate a sign violation risk weight. Conduct relative speed clustering analysis through the set of surrounding vehicle movement trajectories. For example, when it is found that the surrounding vehicles are moving fast and the distance from the target vehicle is small, generate a vehicle distance safety assessment matrix. Then spatially superimpose these visibility attenuation coefficients, road adhesion risk indices, obstacle collision probability distribution maps, sign violation risk weights, and vehicle distance safety assessment matrices. With the target vehicle as the center, divide the spatial area into square grid cells according to a certain radius, calculate and superimpose various risk factors within each grid cell, and finally generate dynamic environmental risk factors.

[0021] When converting in-vehicle communication interaction records into a structured communication feature matrix, for example, the communication interaction between a vehicle and multiple roadside units, record information such as the time, frequency, and data volume of the communication, and organize this information into a matrix form according to certain rules to reflect the communication status and relationship between the vehicle and external devices.

[0022] Step S130, based on the combined operation feature vector, the abnormal behavior pattern identifier, the dynamic environmental risk factor, and the structured communication feature matrix, construct a spatio-temporal association risk analysis map.

[0023] Specifically, the combined operation feature vector can be mapped to the state attribute nodes of the target vehicle. For example, relevant features such as engine speed and braking pressure are mapped to nodes representing the vehicle state. The abnormal behavior pattern identifier is mapped to the driving behavior association edge. For example, a situation with a relatively high risk coefficient of sudden acceleration is mapped to an edge indicating the impact of sudden acceleration risk on the vehicle state. The dynamic environmental risk factor is mapped to the environmental constraint layer. Factors such as reduced visibility and increased road adhesion risk caused by rainy weather constitute environmental constraints on vehicle driving. The structured communication feature matrix is mapped to the communication interaction path, such as the communication path between the vehicle and the roadside unit.

[0024] Perform time slicing division on the state attribute nodes based on a preset time window. For example, take 5 minutes as a time slice, and analyze the changes in the vehicle state within each slice to generate a topological relationship in the time dimension. Perform spatial grid division on the environmental constraint layer according to the vehicle's geographical location data. Assuming the vehicle as the center, divide the area within 100 meters around it into several square spatial grid units to generate a correlation matrix in the spatial dimension. Connect the state attribute nodes within the same time slice through driving behavior association edges. For example, within a time slice, an emergency acceleration behavior association edge connects the motor speed change node and the braking pressure change node to form a driving behavior influence link. Perform cross-layer coupling between the communication interaction path and the driving behavior influence link. For example, after the vehicle communicates with the roadside unit to obtain traffic information, it affects the driver's driving behavior to generate a communication delay risk propagation channel.

[0025] When integrating the above elements, perform timestamp alignment processing on the division results of each time slice in the topological relationship of the time dimension, so that the time order of each time slice is accurate, and generate a time topological node chain with continuous time series attributes. Bind the space-time coordinates of each spatial grid cell in the spatial dimension association matrix to the corresponding time slice in the time topological node chain. For example, associate a certain grid cell with the corresponding vehicle state at a certain moment to generate a space-time encoded grid cell. Extract the start state attribute node and the end state attribute node of each driving behavior association edge in the driving behavior influence link, and map them to the time slice index and spatial grid coordinates of the space-time encoded grid cell. According to the time slice index of the space-time encoded grid cell, perform time window sliding splicing on multiple consecutive driving behavior association edges in the driving behavior influence link to generate a driving behavior propagation path across time slices. Extract the transmission delay parameter of each communication interaction path from the communication delay risk propagation channel, and perform delay compensation offset on the driving behavior propagation path across time slices according to this parameter to generate a driving behavior propagation path after delay calibration. Match the position of each node in the driving behavior propagation path after delay calibration with the spatial grid coordinates of the space-time encoded grid cell to generate a driving behavior propagation path with spatial position constraints. Superimpose the dynamic environmental risk factor of the environmental constraint layer on the spatial grid coordinates of the space-time encoded grid cell to generate the risk density weight of each space-time encoded grid cell. Perform risk conduction probability weighting on the driving behavior propagation path with spatial position constraints according to the risk density weight to generate a risk conduction probability distribution path. Connect the gradient of each node in the risk conduction probability distribution path with the risk density weight of the adjacent space-time encoded grid cell to generate a risk gradient conduction link. Perform time slice backtracking detection on each node in the risk gradient conduction link to identify the set of nodes whose risk density weight exceeds the threshold in consecutive time slices, and generate a risk space-time focus area. According to the spatial grid coordinate distribution of the risk space-time focus area, aggregate multiple nodes located at the same spatial grid coordinate in the risk gradient conduction link to generate a spatially aggregated risk conduction link. Expand each path node in the spatially aggregated risk conduction link in the time dimension, and connect the path nodes of the same spatial grid coordinate in different time slices into a time-evolving risk conduction trajectory. Cross-fuse the time-evolving risk conduction trajectory with the risk gradient conduction link to generate a multi-level risk conduction network with space-time evolution attributes. Dynamically map each node in the multi-level risk conduction network to the risk density weight of the space-time encoded grid cell to generate a node risk weight label. Sort the conduction paths in the multi-level risk conduction network according to the node risk weight label to generate a spatio-temporal association risk conduction map with priority markings.Finally, perform spatial overlay verification on the priority markers in the spatio-temporal correlation risk conduction map and the dynamic environmental risk factors in the environmental constraint layer to eliminate the conduction paths with spatial grid coordinate conflicts. Then, based on the verified spatio-temporal correlation risk conduction map and the time slice division results of the spatio-temporal coding grid cells, generate a spatio-temporal correlation risk analysis map with spatio-temporal stamps.

[0026] Step S140: Dynamically evaluate the driving risks of the target vehicle through the spatio-temporal correlation risk analysis map, and generate a risk level quantification index and a risk event warning label.

[0027] Specifically, traverse the driving behavior impact links in the spatio-temporal correlation risk analysis map and calculate the risk conduction intensity of each link. For example, in a driving behavior impact link, a series of impacts such as sudden changes in motor speed and increased braking pressure caused by rapid acceleration behavior have a relatively high risk conduction intensity. Determine the risk response lag coefficient based on the path length and delay time of the communication delay risk propagation channel. For example, when the communication delay between the vehicle and the roadside unit is relatively long, the risk response lag coefficient is relatively large. Generate an environmental risk superposition effect value based on the spatial grid density and risk factor weight in the environmental constraint layer. In an area with rainfall and obstacles on the road, the environmental risk superposition effect value is relatively high. Monitor the real-time change rate of the eigenvector of the state attribute node to generate a vehicle state degradation index. For example, situations such as too rapid an increase in battery temperature or a sudden increase in tire wear rate will cause the vehicle state degradation index to rise.

[0028] Input the risk conduction intensity, risk response lag coefficient, environmental risk superposition effect value, and vehicle state degradation index into the risk quantification model to output a risk level quantification index. Suppose the risk level quantification index is 80 (out of 100), indicating a relatively high risk. Compare the risk level quantification index with the preset threshold to trigger risk event warning labels at different levels. When the risk level quantification index reaches 80, trigger a high-level risk event warning label, such as "High Risk - Urgent Attention", indicating that the vehicle may face a relatively serious risk situation and timely measures need to be taken.

[0029] Step S150: Generate a risk control decision instruction set based on the risk level quantification index and the risk event warning label, and feedback the risk control decision instruction set to the in-vehicle control terminal and the vehicle networking cloud platform of the target vehicle.

[0030] Specifically, the basic risk control strategy type can be determined according to the risk level quantification index. When the risk level quantification index is 80, it is determined as an emergency braking instruction. Combining with the label type of the risk event warning label "High Risk - Emergency Attention", additional control parameters are matched from the vehicle networking strategy library, such as the parameter for turning on the hazard warning lights simultaneously during emergency braking. Based on the real-time position data of the target vehicle and the high-precision map information, the effective range of the strategy under geographical constraints is generated. For example, obtaining the topological structure data of the road where the target vehicle is currently located shows that there are many curves, the real-time traffic flow information shows that the vehicle density is large, and the high-precision map information shows that the road curvature radius is small, the slope is steep, and it is a two-lane two-way road. Combining with the risk level quantification index, a curve control radius is generated, such as further reducing the safe passing speed of the curve; according to the vehicle density distribution and the risk event warning label, the priority of the avoidance area is generated, such as giving priority to avoiding vehicles in the right lane. Integrating this information defines the spatial boundary and time validity period of the effective range of the strategy, such as within a road range of 200 meters ahead and taking effect within the next 10 minutes.

[0031] By detecting the hardware interface status of the in-vehicle control terminal, the executable instruction set is determined and device conflict instructions are excluded. For example, when it is detected that the braking system is normal, the emergency braking instruction can be executed, while some instructions conflicting with the current operation are excluded. The basic risk control strategy type, additional control parameters, the effective range of the strategy, and the executable instruction set are logically combined to generate a risk control decision instruction set, and then this instruction set is fed back to the in-vehicle control terminal of the target vehicle and the vehicle networking cloud platform. After receiving this instruction set, the vehicle networking cloud platform can comprehensively manage the risk situations of multiple vehicles, such as conducting a global risk situation analysis, generating a regional risk heat map based on the geographical location distribution of multiple target vehicles, adjusting the traffic signal control strategy and the roadside unit broadcast content based on the regional risk heat map, sending the adjusted traffic signal control strategy to the intelligent transportation management system, and updating the roadside unit broadcast content. According to the aggregation result of the risk level quantification index of multiple target vehicles, a macro vehicle networking risk trend report is generated. For example, if the risk level quantification indexes of multiple vehicles in a certain area are relatively high and it is shown as a high-risk area on the regional risk heat map, the traffic signal control strategy in this area is adjusted to extend the red light time to reduce the vehicle flow, and the roadside unit broadcast content is updated to remind drivers to pay attention to the high-risk area ahead and drive carefully.

[0032] Based on the above steps, after the embodiments of the present application collect the multi-dimensional real-time operation data stream of the target vehicle in the vehicle networking environment, key features are respectively extracted from different types of data, including a combined operation feature vector, an abnormal behavior pattern identifier, a dynamic environment risk factor, and a structured communication feature matrix. Then, a spatio-temporal correlation risk analysis map is constructed therefrom, so as to perform deep fusion and correlation analysis on multi-source heterogeneous data, accurately capture the complex spatio-temporal risk relationships during the vehicle operation process, and show the evolution law of the vehicle operation risk in the time and space dimensions. Furthermore, by using the constructed spatio-temporal correlation risk analysis map to dynamically evaluate the driving risk of the target vehicle, a risk level quantification index and a risk event warning label can be generated in real time, and the risk assessment result can be adjusted in a timely manner according to the real-time operation condition of the vehicle, providing timely and accurate risk information for the driver and the relevant management platform, and effectively coping with the continuously changing risk situation during the vehicle driving process. Finally, a risk control decision instruction set is generated according to the risk level quantification index and the risk event warning label, and is fed back to the vehicle-mounted control terminal and the vehicle networking cloud platform. This enables the vehicle networking system to quickly make targeted decisions according to the risk assessment result, realize the active intervention and effective control of the vehicle risk, ensure the safe driving of the vehicle, and improve the safety and reliability of the entire vehicle networking ecosystem.

[0033] In a possible implementation manner, the vehicle state time series data includes an engine speed sequence, a brake pressure waveform, a battery temperature curve, and a tire wear increment, and step S120 includes:

[0034] Step S1201: Perform a frequency domain transformation on the engine speed sequence, and extract a periodic fluctuation amplitude feature and a speed mutation detection mark.

[0035] In this embodiment, for the engine speed sequence in the vehicle state time series data, during the operation of an electric vehicle, the speed of the motor is constantly changing. When performing a frequency domain transformation on the engine speed sequence, it can be understood as analyzing this speed data from another perspective. For example, during the normal acceleration process of the vehicle, the motor speed will increase regularly, and the periodic fluctuation amplitude feature can be extracted through the frequency domain transformation during this process. Assuming that this amplitude feature is in a stable interval under normal conditions, this reflects the stable operation state of the motor under normal working conditions. The speed mutation detection mark is used to detect whether there is an abnormal sudden change in speed during normal driving. For example, when the vehicle is driving on a flat road and suddenly the motor speed drops significantly in an instant, this situation will be recorded by the speed mutation detection mark, which may imply potential fault risks in the motor or the transmission system.

[0036] Step S1202: Perform a segmented integration process on the brake pressure waveform to generate a brake intensity distribution histogram and a pressure abnormal fluctuation threshold.

[0037] In this embodiment, in terms of the braking pressure waveform, every time the driver steps on the brake pedal, the braking system will generate corresponding braking pressure changes. The braking pressure waveform is subjected to piecewise integration processing, and this process is like analyzing the entire braking process in detail according to different stages. For example, during a relatively long braking process, through piecewise integration processing, a braking intensity distribution histogram can be generated. Suppose this histogram shows that most of the braking intensities are concentrated in a relatively low area, which indicates that the driver's braking operations are relatively gentle in most cases. However, if there are some special intervals where the braking intensity suddenly becomes very high, this situation will generate a pressure abnormal fluctuation threshold. For example, during emergency braking, the braking pressure will rise sharply, and the appearance of this pressure abnormal fluctuation threshold can reflect this special braking situation and may also imply abnormalities in the braking system or driving habits.

[0038] Step S1203: Perform trend decomposition on the battery temperature curve to separate the long-term temperature rise gradient feature and the short-term temperature oscillation parameter.

[0039] In this embodiment, the battery temperature curve reflects the temperature change of the battery during vehicle driving. By performing trend decomposition on the battery temperature curve, the internal law of the battery temperature change can be understood more clearly. During a long driving process, the battery temperature will gradually rise, and the long-term temperature rise gradient feature can be separated through trend decomposition. Suppose during a long-distance driving, it is found that the long-term temperature rise gradient feature of the battery temperature shows a relatively steep upward trend, which may mean that there is a problem with the battery's heat dissipation system or the performance of the battery itself is gradually declining. At the same time, during vehicle driving, due to some special working conditions, such as frequent acceleration and deceleration, the battery temperature will have some small rapid fluctuations, and these fluctuations can obtain short-term temperature oscillation parameters through trend decomposition. For example, when the vehicle frequently starts and stops on urban roads, the battery temperature will have some short-term up and down fluctuations, and this fluctuation situation is recorded by the short-term temperature oscillation parameters.

[0040] Step S1204: Perform wear rate modeling on the tire wear increment to calculate the tire life prediction index and the probability of abnormal tread wear.

[0041] In this embodiment, the tire wear increment accumulates gradually as the vehicle mileage increases. Modeling the wear rate of the tire wear increment can be understood as constructing a mathematical model to accurately describe the speed of tire wear. The calculated tire life prediction index can be used to estimate how many more miles the tire can be used. For example, if the tire life prediction index shows that the tire only has 20% of its remaining life, this prompts the driver to pay attention to the condition of the tire and consider whether to replace the tire. The probability of abnormal tread wear reflects the possibility of abnormal wear of the tire. Suppose the vehicle often travels on roads with poor road conditions, such as roads full of potholes and gravel, the probability of abnormal tread wear will be relatively high, which may be due to the uneven road surface causing increased tire wear.

[0042] Step S1205: Normalize and splice the periodic fluctuation amplitude feature, the rotational speed mutation detection mark, the braking intensity distribution histogram, the pressure abnormal fluctuation threshold, the long-term temperature rise gradient feature, the short-term temperature oscillation parameter, the tire life prediction index, and the probability of abnormal tread wear to form the combined operation feature vector.

[0043] In this embodiment, the combined operation feature vector is a comprehensive index set that comprehensively reflects various state characteristics of the vehicle during operation. For example, the numerical values of each element in the combined operation feature vector can be input into the subsequent analysis system for overall evaluation of the vehicle's operation health status.

[0044] In a possible implementation manner, step S120 further includes:

[0045] Step S1206: Screen the number of times of sudden acceleration event triggers, the time interval of consecutive lane change events, the proportion of overspeed duration, and the frequency of illegal use of lights at night from the driving behavior event set.

[0046] Step S1207: Generate a sudden acceleration risk coefficient based on the comparison between the number of times of sudden acceleration event triggers and a preset acceleration threshold.

[0047] For example, during the driving process of this electric vehicle, the number of times of sudden acceleration event triggers is an important indicator. Suppose that during a specific driving distance, the sudden acceleration event is triggered 3 times. Comparing this number of times of sudden acceleration event triggers with the preset acceleration threshold can generate a sudden acceleration risk coefficient. If the preset acceleration threshold is a certain value and the acceleration in these 3 sudden acceleration events exceeds this threshold, then the sudden acceleration risk coefficient will be relatively high, indicating that the driver's sudden acceleration operation may pose a greater risk to the safety and performance of the vehicle.

[0048] Step S1208: Determine the intensity score of lane-changing behavior based on the statistical distribution of the time intervals of the continuous lane-changing events.

[0049] On multi-lane urban roads, drivers may perform continuous lane-changing operations. By statistically analyzing the time intervals of continuous lane-changing events, the intensity score of lane-changing behavior can be determined. Suppose on a certain section of road, a driver makes 3 consecutive lane-changes within a short distance, and the time intervals between each lane-change are very short, which will result in a relatively high intensity score of lane-changing behavior. This indicates that the driver's lane-changing behavior is relatively frequent and hasty, increasing the risk of collision with other vehicles.

[0050] Step S1209: Generate the cumulative amount of speeding violations by combining the proportion of speeding duration and the road speed limit data.

[0051] During the entire driving process, the proportion of speeding duration is calculated in combination with the road speed limit data. For example, on a road with a speed limit of 60 km / h, the total driving time of the vehicle is 30 minutes, and the time of speeding (speed exceeding 60 km / h) is 10 minutes. Then the proportion of speeding duration is 10 / 30 ≈ 33.3%. By combining this proportion of speeding duration with the road speed limit data, the cumulative amount of speeding violations can be generated. If this cumulative amount of speeding violations is relatively large, it indicates that the driver has relatively serious speeding violations on this section of the road, increasing the risk of traffic accidents.

[0052] Step S1210: Generate a lighting violation pattern identifier by matching the frequency of nighttime illegal lighting use with the lighting specification database.

[0053] When driving at night, there are clear specifications for the use of vehicle lights. By statistically analyzing the frequency of nighttime illegal lighting use and then matching it with the lighting specification database, a lighting violation pattern identifier can be generated. For example, if a driver often fails to use high beams or turn signals in accordance with regulations at night, the frequency of nighttime illegal lighting use will be relatively high, and the generated lighting violation pattern identifier will show this illegal behavior pattern.

[0054] Step S1211: Perform weighted fusion on the rapid acceleration risk coefficient, the intensity score of lane-changing behavior, the cumulative amount of speeding violations, and the lighting violation pattern identifier to generate the abnormal behavior pattern identifier.

[0055] In this embodiment, the abnormal behavior pattern identifier can be understood as a comprehensive evaluation result, which comprehensively reflects whether there is an abnormal risk in the driver's driving behavior. For example, if the rapid acceleration risk coefficient is high, the density score of lane-changing behavior is high, the cumulative amount of speeding violations is large, and the lighting violation pattern identifier shows many violations, then the abnormal behavior pattern identifier after weighted fusion will show a high risk level, indicating that the driver's driving behavior needs to be corrected or monitored to ensure the safety of vehicle driving.

[0056] In a possible implementation manner, step S120 further includes:

[0057] Step S1212, identifying real-time meteorological condition parameters, a heat map of road obstacle distribution, traffic sign semantic information, and a set of surrounding vehicle movement trajectories from the external environment perception data.

[0058] Step S1213, calculating a visibility attenuation coefficient and a road surface adhesion risk index according to the real-time meteorological condition parameters.

[0059] Step S1214, generating an obstacle collision probability distribution map based on the density gradient of the road obstacle distribution heat map.

[0060] Step S1215, performing compliance verification on the traffic sign semantic information to generate a sign violation risk weight.

[0061] Step S1216, performing relative speed clustering analysis through the set of surrounding vehicle movement trajectories to generate a vehicle spacing safety evaluation matrix.

[0062] For real-time meteorological condition parameters, during the driving process of an electric vehicle, sensors continuously monitor the meteorological conditions. For example, when encountering rainy weather, the meteorological sensor will detect parameters such as rainfall intensity and humidity, and calculate the visibility attenuation coefficient and the road surface adhesion risk index according to these real-time meteorological condition parameters. Rainfall will cause a decrease in visibility, and the visibility attenuation coefficient is calculated through an algorithm. Suppose in a moderate rain weather, the visibility decreases from several kilometers under normal conditions to a few hundred meters, and this reduction ratio and degree are reflected in the visibility attenuation coefficient. At the same time, rainfall will make the road surface wet and slippery, and different road surface materials have different adhesion capabilities in the wet state. According to this situation, the road surface adhesion risk index is calculated. For example, for an asphalt road surface, the road surface adhesion risk index will increase significantly after rainfall, indicating that the vehicle is more likely to skid when braking and steering.

[0063] On urban roads, there are many other vehicles around, and the vehicle's sensors will track the movement trajectories of these surrounding vehicles. Through relative speed clustering analysis of the set of movement trajectories of surrounding vehicles, for example, vehicles with similar speeds and relatively stable spacings are grouped into one category. If it is found that the surrounding vehicles are moving fast and the spacing from the target electric vehicle is small, corresponding evaluation values will be generated in the vehicle spacing safety evaluation matrix. Suppose on a multi-lane road, the vehicle in the adjacent lane approaches the target vehicle quickly and the spacing is continuously shrinking, and the evaluation values in the vehicle spacing safety evaluation matrix will show a high risk level.

[0064] Step S1217, spatially overlay the visibility attenuation coefficient, the road surface adhesion risk index, the obstacle collision probability distribution map, the sign violation risk weight, and the vehicle spacing safety evaluation matrix to generate the dynamic environment risk factor.

[0065] For example, step S1217 includes:

[0066] Step S1217-1, based on the geographical coordinate information in the real-time meteorological condition parameters and the coverage range of the road obstacle distribution heat map, divide the spatial area within a preset radius around the target vehicle into square grid cells of equal size to generate an environmental spatial overlay grid.

[0067] For example, with the target electric vehicle as the center, within a range of 500 meters in radius, it is divided into square grid cells with a side length of 10 meters.

[0068] Step S1217-2, perform spatial interpolation calculation on the visibility attenuation coefficient according to the center position of each grid cell of the environmental spatial overlay grid to generate a visibility attenuation coefficient distribution raster corresponding to each grid cell.

[0069] Suppose at the center position of a certain grid cell, according to the relationship between its distance from the surrounding meteorological sensor detection points and the visibility attenuation coefficient, the visibility attenuation coefficient distribution raster value of this grid cell is obtained through spatial interpolation calculation. If this grid cell is close to the rainfall center area, the visibility attenuation coefficient distribution raster value will be higher.

[0070] Step S1217-3, according to the preset mapping relationship of the road surface adhesion risk index in different road material areas, match and fill the road surface adhesion risk index according to the road material type to which the grid cells of the environmental spatial overlay grid belong to generate a road surface adhesion risk index distribution raster.

[0071] For example, the road area covered by a certain grid cell is a cement road surface. According to the preset value of the road surface adhesion risk index of the cement road surface after rainfall, this value is filled into the road surface adhesion risk index distribution raster of this grid cell.

[0072] Step S1217-4: Resample the probability density of the obstacle collision probability distribution map according to the grid cell boundaries of the environmental space overlay grid, and generate a normalized grid of obstacle collision probabilities for each grid cell.

[0073] If there is a partial area in a certain grid cell close to a broken-down vehicle on the road, through probability density resampling, the value of the normalized grid of obstacle collision probabilities for this grid cell will be relatively high, indicating a relatively high possibility of a vehicle colliding with an obstacle within this grid cell.

[0074] Step S1217-5: According to the geographical location data of the traffic sign semantic information, project the sign violation risk weight into the corresponding grid cells of the environmental space overlay grid to generate a sign violation risk weight distribution grid.

[0075] For example, in a grid cell close to a speed limit sign, due to the risk of a vehicle speeding, after the sign violation risk weight is projected into this grid cell, the value of the sign violation risk weight distribution grid will be relatively high.

[0076] Step S1217-6: Extract the relative position coordinates of each surrounding vehicle with respect to the target vehicle in the vehicle spacing safety assessment matrix, and perform nearest neighbor assignment on the assessment values of the vehicle spacing safety assessment matrix according to the grid cells of the environmental space overlay grid to generate a vehicle spacing safety assessment distribution grid.

[0077] If the distance between a certain surrounding vehicle and the target vehicle is small and it is near a certain grid cell, then the value of the vehicle spacing safety assessment distribution grid for this grid cell will reflect this close-range risk situation.

[0078] Step S1217-7: Perform unified normalization processing on the visibility attenuation coefficient distribution grid, the road surface adhesion risk index distribution grid, the normalized grid of obstacle collision probabilities, the sign violation risk weight distribution grid, and the vehicle spacing safety assessment distribution grid, so that the numerical ranges of each grid are mapped to the same risk level interval.

[0079] For example, convert the value of the visibility attenuation coefficient distribution grid from the original actual measurement value range to a value between 0 and 1, and perform similar conversions on other grids to ensure that they are under the same risk level evaluation system.

[0080] Step S1217-8: Based on the preset visibility risk weight, road surface adhesion risk weight, obstacle collision risk weight, sign violation risk weight, and vehicle spacing risk weight, perform weighted summation on the normalized distribution grids according to the corresponding grid cells to generate a comprehensive risk superposition value for each grid cell.

[0081] Assume that the visibility risk weight is 0.2, the road surface adhesion risk weight is 0.3, the obstacle collision risk weight is 0.2, the sign violation risk weight is 0.1, and the vehicle spacing risk weight is 0.2. In a certain grid cell, the comprehensive risk superposition value is calculated through weighted summation.

[0082] Step S1217-9: According to the distribution of the comprehensive risk superposition value, perform spatial smoothing filtering on the comprehensive risk superposition value of each grid cell in the environmental space superposition grid to eliminate the noise interference of isolated high-risk grid cells.

[0083] For example, due to a short-term error of the sensor, a certain grid cell may have an extremely high comprehensive risk superposition value, but the risk values of the surrounding grid cells are relatively low. Through spatial smoothing filtering, the risk value of this isolated high-risk grid cell can be reduced to make it more in line with the actual risk distribution.

[0084] Step S1217-10: Segment and mark the comprehensive risk superposition value after spatial smoothing filtering according to the preset risk level division threshold to generate a dynamic environmental risk factor spatial distribution map including multi-level risk regions.

[0085] For example, the comprehensive risk superposition value is divided into three levels: low risk (0-0.3), medium risk (0.3-0.6), and high risk (0.6-1). In the dynamic environmental risk factor spatial distribution map, regions of different levels are marked with different colors or identifiers.

[0086] Step S1217-11: According to the real-time movement direction and speed vector of the target vehicle, perform risk propagation rate compensation on the grid cells within a preset distance in front of the vehicle in the dynamic environmental risk factor spatial distribution map to generate a dynamic environmental risk factor extended in the time dimension.

[0087] Assume that the target electric vehicle is moving forward at a certain speed. For the grid cells within 200 meters in front of the vehicle, considering that the vehicle will enter these regions in the future period of time, according to the vehicle's speed and driving direction, the risk values of these grid cells are adjusted to reflect the propagation of risk in the time dimension. For example, if there is a high-risk region in front of the vehicle and the vehicle is approaching quickly, then after the risk propagation rate compensation, the risk value of this high-risk region will further increase, so as to more accurately reflect the risk situation that the vehicle is about to face.

[0088] In a possible implementation manner, step S130 includes:

[0089] Step S131: Map the combined operation feature vector to the state attribute node of the target vehicle, map the abnormal behavior pattern identifier to the driving behavior association edge, map the dynamic environment risk factor to the environmental constraint layer, and map the structured communication feature matrix to the communication interaction path.

[0090] First, map the combined operation feature vector to the state attribute node of the target vehicle. For example, elements in the combined operation feature vector such as the periodic fluctuation amplitude feature in the engine speed sequence and the rotational speed mutation detection mark can be mapped to nodes representing the vehicle state. These nodes reflect various aspects of the internal operation state of the vehicle. The abnormal behavior pattern identifier is mapped to the driving behavior association edge. The abnormal behavior pattern identifier composed of factors such as the rapid acceleration risk coefficient and the lane change behavior density score can construct an edge representing the driving behavior association, such as the association edge between the rapid acceleration behavior and the vehicle state change. The dynamic environment risk factor is mapped to the environmental constraint layer. For example, the dynamic environment risk factor composed of the visibility attenuation coefficient, the road surface adhesion risk index, etc. constitutes the environmental constraint for vehicle driving. For example, the reduced visibility and slippery road surface caused by rainfall are an environmental constraint. The structured communication feature matrix is mapped to the communication interaction path. For example, the structured communication feature matrix composed of the communication interaction records between the vehicle and the roadside unit or other vehicles reflects the communication path relationship between the vehicle and the external world.

[0091] Step S132: Based on a preset time window, perform time slicing on the state attribute nodes to generate a topological relationship in the time dimension.

[0092] Suppose the preset time window is 5 minutes. Analyze the vehicle state attribute nodes within each 5 - minute time period. For example, the state attribute nodes such as the engine speed and braking pressure within 0 - 5 minutes form a time slice, and the state attribute nodes within 5 - 10 minutes form another time slice. The order and relationship between these time slices constitute the topological relationship in the time dimension.

[0093] Step S133: According to the vehicle geographical location data, perform spatial grid division on the environmental constraint layer to generate a correlation matrix in the spatial dimension.

[0094] For example, with the current vehicle position as the center, divide the surrounding area into square spatial grid units. For example, divide the area of 100 meters × 100 meters around the vehicle into square grid units with a side length of 10 meters. Each grid unit corresponds to different environmental constraint information, and the relationship between these grid units forms a correlation matrix in the spatial dimension.

[0095] Step S134: Connect the state attribute nodes within the same time slice through the driving behavior association edge to form a driving behavior influence link.

[0096] For example, within the same 5-minute time slice, the sharp acceleration behavior correlation edge may connect the state attribute node with a sudden change in engine speed and the state attribute node with a change in braking pressure. In this way, a driving behavior influence link is constructed, indicating the influence path of driving behavior on vehicle state during this time period.

[0097] Step S135: Cross-layer couple the communication interaction path and the driving behavior influence link to generate a communication delay risk propagation channel.

[0098] For example, when a vehicle communicates with a roadside unit to obtain traffic congestion information, this communication interaction path is coupled with the driving behavior influence link where the driver changes driving behavior (such as decelerating or changing lanes) due to traffic congestion. Since there may be a delay in communication, this delay will affect the adjustment of driving behavior, thus forming a communication delay risk propagation channel.

[0099] Step S136: Integrate the time dimension topological relationship, the space dimension association matrix, the driving behavior influence link, and the communication delay risk propagation channel to form the spatio-temporal association risk analysis map.

[0100] For example, step S136 includes:

[0101] Step S1361: Perform timestamp alignment processing on the division results of each time slice in the time dimension topological relationship to generate a time topological node chain with continuous time series attributes.

[0102] Step S1362: Bind the spatio-temporal coordinates of each spatial grid cell in the space dimension association matrix to the corresponding time slice in the time topological node chain to generate spatio-temporal encoded grid cells.

[0103] Step S1363: Extract the starting state attribute node and the ending state attribute node of each driving behavior correlation edge in the driving behavior influence link, and map the starting state attribute node and the ending state attribute node to the time slice index and spatial grid coordinates of the spatio-temporal encoded grid cell.

[0104] Step S1364: According to the time slice index of the spatio-temporal encoded grid cell, perform time window sliding splicing on consecutive multiple driving behavior correlation edges in the driving behavior influence link to generate a driving behavior propagation path across time slices.

[0105] Step S1365: Extract the transmission delay parameter of each communication interaction path from the communication delay risk propagation channel, and perform delay compensation offset on the driving behavior propagation path across time slices according to the transmission delay parameter to generate a driving behavior propagation path with delay calibration.

[0106] Step S1366: Match the position of each node in the driving behavior propagation path after latency calibration with the spatial grid coordinates of the spatio-temporal coding grid cells to generate a driving behavior propagation path with spatial position constraints.

[0107] Step S1367: Superimpose the dynamic environmental risk factors of the environmental constraint layer on the spatial grid coordinates of the spatio-temporal coding grid cells to generate the risk density weight of each spatio-temporal coding grid cell.

[0108] Step S1368: Perform risk conduction probability weighting on the driving behavior propagation path with spatial position constraints according to the risk density weight to generate a risk conduction probability distribution path.

[0109] Step S1369: Connect the gradient of each node in the risk conduction probability distribution path with the risk density weight of adjacent spatio-temporal coding grid cells to generate a risk gradient conduction link.

[0110] Step S13610: Perform time-slice backtracking detection on each node in the risk gradient conduction link, identify the set of nodes whose risk density weight exceeds the threshold in consecutive time slices, and generate a risk spatio-temporal focus area.

[0111] Step S13611: Aggregate the paths of multiple nodes located at the same spatial grid coordinate in the risk gradient conduction link according to the spatial grid coordinate distribution of the risk spatio-temporal focus area to generate a spatially aggregated risk conduction link.

[0112] Step S13612: Expand each path node in the spatially aggregated risk conduction link in the time dimension, and connect the path nodes at the same spatial grid coordinate in different time slices into a time-evolving risk conduction trajectory.

[0113] Step S13613: Cross-fuse the time-evolving risk conduction trajectory with the risk gradient conduction link to generate a multi-level risk conduction network with spatio-temporal evolution attributes.

[0114] Step S13614: Dynamically map each node in the multi-level risk conduction network with the risk density weight of the spatio-temporal coding grid cells to generate a node risk weight label.

[0115] Step S13615: Sort the conduction paths in the multi-level risk conduction network according to the node risk weight label to generate a spatio-temporal associated risk conduction map with priority markings.

[0116] Step S13616: Perform spatial overlay verification on the priority markings in the spatio-temporal correlation risk conduction map and the dynamic environmental risk factors in the environmental constraint layer, eliminate the conduction paths with spatial grid coordinate conflicts, and generate the spatio-temporal correlation risk analysis map with spatio-temporal stamps based on the verified spatio-temporal correlation risk conduction map and the time slice division result of the spatio-temporal coding grid cell.

[0117] In this embodiment, time stamp alignment processing is performed on each time slice division result in the time dimension topological relationship to generate a time topological node chain with continuous time series attributes. Ensure that the start and end times of each time slice are accurately marked, so that each time slice is connected in sequence to form a continuous time topological node chain. For example, time slices such as 0 - 5 minutes, 5 - 10 minutes are accurately connected. Bind each spatial grid cell in the spatial dimension correlation matrix to the corresponding time slice in the time topological node chain to generate a spatio-temporal coding grid cell. For example, within a specific 5-minute time slice, the spatial grid cell at a specific position around the vehicle is bound to it to form a spatio-temporal coding grid cell with spatio-temporal attributes.

[0118] Extract the start state attribute node and the end state attribute node of each driving behavior association edge in the driving behavior influence link, and map the start state attribute node and the end state attribute node to the time slice index and spatial grid coordinates of the spatio-temporal coding grid cell. For example, in a certain driving behavior association edge, the start state attribute node is the engine speed node during hard acceleration (located at a specific time slice index), and the end state attribute node is the brake pressure change node. Map these two nodes to the time slice index and spatial grid coordinates of the corresponding spatio-temporal coding grid cell. According to the time slice index of the spatio-temporal coding grid cell, perform time window sliding splicing on multiple consecutive driving behavior association edges in the driving behavior influence link to generate a driving behavior propagation path across time slices. For example, the driving behavior association edges within multiple consecutive 5-minute time slices are spliced in sequence to form a driving behavior propagation path across time, reflecting the propagation of driving behavior over a longer time span.

[0119] Extract the transmission delay parameters of each communication interaction path from the communication delay risk propagation channel, and perform delay compensation offset on the driving behavior propagation path across time slices according to the transmission delay parameters to generate a driving behavior propagation path with delay calibration. Assume that the transmission delay of a certain communication interaction path is 2 seconds. When constructing the driving behavior propagation path, take this delay into account and perform temporal offset adjustment on the relevant driving behavior association edges to make the driving behavior propagation path more in line with the actual situation. Match the position of each node in the driving behavior propagation path with the spatial grid coordinates of the spatio-temporal coding grid unit to generate a driving behavior propagation path with spatial position constraints. In this way, each node in the driving behavior propagation path corresponds to the spatial position around the vehicle, clarifying the propagation of driving behavior in space.

[0120] Overlay the dynamic environmental risk factors of the environmental constraint layer on the spatial grid coordinates of the spatio-temporal coding grid unit to generate the risk density weight of each spatio-temporal coding grid unit. For example, in a certain spatio-temporal coding grid unit, if the visibility attenuation coefficient in this area is high and the road surface adhesion risk index is large, then the risk density weight of this grid unit will be high. Perform risk conduction probability weighting on the driving behavior propagation path with spatial position constraints according to the risk density weight to generate a risk conduction probability distribution path. If the driving behavior propagation path passes through a spatio-temporal coding grid unit with a high risk density weight, then the risk conduction probability in this area will increase accordingly, thus forming a risk conduction probability distribution path.

[0121] Connect the gradient of the risk density weight of each node in the risk conduction probability distribution path with that of the adjacent spatio-temporal coding grid unit to generate a risk gradient conduction link. For example, the risk conduction probability of a certain node is affected by the risk density weight of the adjacent grid unit, and the difference in the risk density weight of the adjacent grid unit forms the gradient of risk conduction. In this way, a risk gradient conduction link is constructed. Perform time slice backtracking detection on each node in the risk gradient conduction link to identify the set of nodes whose risk density weight exceeds the threshold in consecutive time slices, and generate a risk spatio-temporal focus area. If within multiple consecutive time slices, the risk density weight of the spatio-temporal coding grid unit where a certain node is located is always high and exceeds the set threshold, this node is included in the risk spatio-temporal focus area, indicating that this area is a high-risk area in both time and space.

[0122] According to the spatial grid coordinate distribution of the risk spatio-temporal focusing area, multiple nodes located at the same spatial grid coordinate in the risk gradient conduction link are path-aggregated to generate a spatial aggregated risk conduction link. For example, at the same spatial grid coordinate, risk-related nodes in different time slices are aggregated to form a spatial aggregated risk conduction link, reflecting the comprehensive situation of risk conduction at this spatial position. Each path node in the spatial aggregated risk conduction link is extended in the time dimension, and the path nodes of the same spatial grid coordinate in different time slices are connected into a time-evolving risk conduction trajectory. In this way, the evolution of risk over time at the same spatial position can be seen.

[0123] The time-evolving risk conduction trajectory is cross-fused with the risk gradient conduction link to generate a multi-level risk conduction network with spatio-temporal evolution attributes. This multi-level risk conduction network contains risk conduction relationships at different time and space levels, comprehensively reflecting the spread of risks. Each node in the multi-level risk conduction network is dynamically mapped with the risk density weight of the spatio-temporal coding grid unit to generate a node risk weight label. According to the risk density weight situation of the node in different spatio-temporal coding grid units, a risk weight label is attached to each node. The conduction paths in the multi-level risk conduction network are sorted by risk intensity according to the node risk weight labels to generate a spatio-temporal associated risk conduction map with priority markings. The conduction path where the node with a higher risk weight is located has a higher priority.

[0124] Finally, the priority markings in the spatio-temporal associated risk conduction map are spatially superimposed and verified with the dynamic environmental risk factors in the environmental constraint layer to eliminate the conduction paths with spatial grid coordinate conflicts. For example, if there is a conflict between the priority marking of a certain conduction path and the actual risk situation in the environmental constraint layer (such as a certain area is actually of low risk but the conduction path shows high priority), this conduction path is adjusted. According to the verified spatio-temporal associated risk conduction map and the time slice division result of the spatio-temporal coding grid unit, a spatio-temporal associated risk analysis map with spatio-temporal stamps is generated, which accurately reflects the risk situation of the vehicle in the spatio-temporal dimension.

[0125] In a possible implementation manner, step S140 includes:

[0126] Step S141, traversing the driving behavior impact link in the spatio-temporal associated risk analysis map and calculating the risk conduction intensity of each link.

[0127] For example, in a driving behavior impact chain, starting from the hard acceleration behavior, passing through nodes such as sudden engine speed change and brake pressure change, the risk conduction relationship between each node has a certain intensity. Hard acceleration may cause the engine to suddenly bear a large load. When this load change is transmitted to the braking system, if the brake pressure cannot respond in a timely and effective manner, the risk will increase. By analyzing factors such as the correlation between each node in the chain, the numerical values of relevant eigenvectors, and environmental constraints, the risk conduction intensity of this chain is quantified. For example, when the amplitude of the sudden engine speed change is large, the brake pressure response is lagging, and the environmental constraint is a slippery road surface (increasing the risk of braking distance), the risk conduction intensity of this chain will be relatively high.

[0128] Step S142: Determine the risk response lag coefficient according to the path length and delay time of the communication delay risk propagation channel.

[0129] In the scenario of vehicle communication with roadside units or other vehicles, the path length of the communication delay risk propagation channel depends on factors such as the distance between communication devices and the topology of the communication network. For example, when a vehicle communicates with a roadside unit at a relatively long distance, the path is longer. At the same time, there may be a communication delay time. Suppose there is a 2-second delay in the vehicle obtaining traffic congestion information. A longer path length and a larger delay time will cause the risk response lag coefficient to increase. This means that the driver will react later based on the information obtained through communication, increasing the driving risk. For example, when the vehicle receives information that there is a sudden accident ahead and needs to decelerate, if the driver starts braking 2 seconds later due to communication delay, the vehicle may have approached the accident area during this period, greatly increasing the risk.

[0130] Step S143: Generate an environmental risk superposition effect value based on the spatial grid density and risk factor weights of the environmental constraint layer.

[0131] The environmental constraint layer consists of multiple factors, such as the visibility attenuation coefficient and the road surface adhesion risk index. Taking the division of the area around the vehicle into multiple spatial grid units as an example, there may be a relatively high spatial grid density in some areas, that is, the area where environmental constraint factors are concentrated. For example, near an intersection, there are many traffic signs, a large vehicle density, and there may be road obstacles, so the spatial grid density in this area is relatively high. At the same time, each risk factor (such as the risk factor weight corresponding to the visibility attenuation coefficient, the risk factor weight corresponding to the road surface adhesion risk index, etc.) has a corresponding weight according to its impact on the risk. By comprehensively considering the spatial grid density and the weights of each risk factor, the environmental risk superposition effect value is calculated. In the example of the intersection, due to the concentration of multiple risk factors and large weights, the environmental risk superposition effect value will be relatively high.

[0132] Step S144, monitor the real-time change rate of the feature vector of the state attribute node to generate a vehicle state degradation index.

[0133] The state attribute node includes relevant feature vectors such as the engine speed sequence and the brake pressure waveform. For example, for the engine speed sequence, its change rate is monitored in real time. If during normal driving, the engine speed suddenly shows frequent large fluctuations, or the fluctuation amplitude gradually increases over time, this indicates that the operating state of the engine may be degrading. For the brake pressure waveform, if the response time of the brake pressure gradually becomes longer or the fluctuation of the brake pressure becomes unstable, it also indicates that there may be a problem with the braking system. By comprehensively analyzing the real-time change rate of the feature vectors of the above state attribute nodes, a vehicle state degradation index is generated. For example, abnormal changes in the engine speed and the brake pressure waveform result in an increase in the vehicle state degradation index, indicating that the operating health status of the vehicle's internal system is declining, increasing the driving risk.

[0134] Step S145, input the risk conduction intensity, the risk response lag coefficient, the environmental risk superposition effect value, and the vehicle state degradation index into the risk quantification model, and output the risk level quantification index.

[0135] The risk quantification model is a tool for accurately assessing risks constructed based on historical data. The risk conduction intensity reflects the propagation risk of driving behavior in the vehicle's internal system and the environment. The risk response lag coefficient reflects the impact of external communication on driving risks. The environmental risk superposition effect value represents the comprehensive risk impact of the external environment on the vehicle. The vehicle state degradation index represents the health status risk of the vehicle's own system. After these factors are jointly input into the risk quantification model, the model outputs a risk level quantification index according to its internal algorithms and parameters. For example, after model calculation, the risk level quantification index is 70 (assuming a full score of 100), and this value indicates that the vehicle is currently at a relatively high risk level.

[0136] Step S146, compare the risk level quantification index with a preset threshold to trigger risk event warning labels at different levels.

[0137] If the preset threshold is 60, when the risk level quantification index is 70, exceeding the preset threshold will trigger a risk event warning label at the corresponding level. For example, it may trigger a risk event warning label of "medium-high risk - need to pay close attention", prompting the driver of the vehicle or the relevant monitoring system that the vehicle is currently facing a medium-high degree of risk and requires close attention to the vehicle state, driving behavior, or environmental conditions, or taking corresponding measures.

[0138] In a possible implementation manner, the construction steps of the risk quantification model include:

[0139] Step S210: Obtain the sample set of risk events marked in the historical vehicle networking data, and extract the combined operation feature vectors, abnormal behavior pattern identifiers, dynamic environmental risk factors, and structured communication feature matrices corresponding to the samples.

[0140] In the historical driving data of electric vehicles, there is a large amount of data that has been marked as risk events or normal events. For example, collision accidents caused by sudden acceleration that have occurred in the past, and rubbing events that occurred due to low visibility in bad weather are marked as risk events, while the data during normal driving is marked as normal events. Extract the combined operation feature vectors, abnormal behavior pattern identifiers, dynamic environmental risk factors, and structured communication feature matrices corresponding to the samples from these samples. For risk event samples, the combined operation feature vectors may include features such as abnormal fluctuations in engine speed and unstable braking pressure; the abnormal behavior pattern identifiers may show a relatively high number of sudden acceleration event triggers and a high density of continuous lane-changing behaviors; the dynamic environmental risk factors may include factors such as low visibility and slippery road surface in rainy weather; the structured communication feature matrices may reflect situations such as communication delays or interruptions before an accident occurs.

[0141] Step S220: Construct an initial risk assessment neural network, where the initial risk assessment neural network includes a feature fusion layer, a time attention layer, and a spatial convolution layer.

[0142] Step S230: Perform cross-modal fusion on the combined operation feature vectors and the dynamic environmental risk factors through the feature fusion layer to generate environmental coupling features.

[0143] The initial risk assessment neural network is a complex model constructed to accurately assess driving risks. The role of the feature fusion layer is to perform cross-modal fusion on the combined operation feature vectors and the dynamic environmental risk factors to generate environmental coupling features. For example, fuse the internal operation features of the vehicle such as engine speed and braking pressure with the external environmental features such as visibility and road adhesion. In this process, features of different modalities (internal vehicle state and external environment) are fused through specific algorithms to generate environmental coupling features that can comprehensively reflect the state of the vehicle in the environment.

[0144] Step S240: Perform time-dependence modeling on the abnormal behavior pattern identifiers through the time attention layer to generate behavior time series weights.

[0145] In historical data, factors such as the number of hard acceleration event triggers and the time intervals of consecutive lane change events in the abnormal behavior pattern identifier have different importance at different time points. For example, during traffic congestion periods, hard acceleration events may have a higher risk weight than during normal driving periods. The time attention layer analyzes the variation patterns of these abnormal behavior patterns in the time series, assigns corresponding behavioral time series weights to the abnormal behavior patterns at each time point to reflect their risk importance at different times.

[0146] Step S250, perform spatial feature extraction on the structured communication feature matrix through the spatial convolution layer to generate a communication risk distribution map.

[0147] The structured communication feature matrix contains spatial relationship information of vehicle communication with surrounding devices. Through the convolution operation of the spatial convolution layer, spatial features of communication can be extracted, such as the distribution of communication signal strength at different spatial positions, information on areas where communication interruptions may occur, etc., thereby generating a communication risk distribution map, which intuitively shows the spatial distribution of communication risks.

[0148] Step S260, perform a fully connected process on the environmental coupling feature, the behavioral time series weight, and the communication risk distribution map, and output a risk prediction value.

[0149] The fully connected layer comprehensively calculates the previously processed different features, and based on the internal connection weights and activation functions, outputs a risk prediction value, which is a preliminary risk estimate based on the model's comprehensive evaluation of the input features.

[0150] Step S270, based on the error backpropagation between the risk prediction value and the actual risk annotation value, optimize the parameters of the initial risk assessment neural network to obtain the risk quantification model.

[0151] During the training process of the initial risk assessment neural network, there is an error between the risk prediction value and the actual risk annotation value (annotations of risk events or normal events known from historical data). Through the error backpropagation algorithm, the error is propagated backward from the output layer to the previous layers, adjusting the parameters of each layer (such as the fusion weights of the feature fusion layer, the time weights of the time attention layer, the convolution kernel parameters of the spatial convolution layer, etc.), continuously optimizing the model, so that the error between the risk prediction value and the actual risk annotation value gradually decreases, and finally an accurate risk quantification model is obtained.

[0152] In a possible implementation manner, step S150 includes:

[0153] Step S151, determine the basic risk control strategy type according to the risk level quantification index, including an emergency braking instruction, a speed limit control instruction, a path replanning instruction, or a driver warning instruction.

[0154] Specifically, when the risk level quantification index reaches a certain value, corresponding basic risk control strategy types need to be adopted. For example, if the risk level quantification index is high, indicating that the vehicle faces a greater risk, an emergency braking command may need to be issued. During the driving of an electric vehicle, if an obstacle suddenly appears ahead and the risk level quantification index exceeds the threshold of emergency braking, it will be determined as the type of emergency braking command to avoid the collision risk. If the risk level quantification index is at a medium level, it may be a speed limit control command. For example, when the road conditions are complex or the vehicle performance shows certain abnormalities (such as too high battery temperature), in order to ensure driving safety and the stability of the vehicle system, the vehicle speed is limited within a reasonable range. When the risk level quantification index indicates that there are many risk factors in the surrounding environment (such as road construction, traffic congestion), a route replanning command may be triggered to guide the vehicle to select a safer and smoother route. And when the risk level quantification index does not reach the emergency level but there are some potential risk factors (such as the initial signs of driver fatigue driving), a driver warning command will be issued to remind the driver to pay attention to safe driving.

[0155] Step S152: Combine the label type of the risk event warning label and match additional control parameters from the vehicle networking policy library.

[0156] The risk event warning label contains more detailed risk information. For example, if the risk event warning label is "low visibility ahead - high risk", combine this label type and match additional control parameters from the vehicle networking policy library. For the emergency braking command, parameters such as turning on the hazard warning lights and increasing the braking assistance force during braking may be matched; for the speed limit control command, parameters such as further reducing the speed limit value according to the visibility situation may be matched. For example, on a road with a normal speed limit of 60 km / h, due to low visibility ahead, the speed limit value is reduced to 30 km / h; for the route replanning command, parameters such as preferentially selecting roads with street lights or clear traffic signs as the new route will be matched; for the driver warning command, additional control parameters such as simultaneously reminding the driver to pay attention to the low visibility ahead, drive carefully, and turn on the fog lights through voice and screen display will be matched.

[0157] Step S153: Generate the effective range of the strategy under geographical constraints based on the real-time position data of the target vehicle and the high-precision map information.

[0158] Step S154: Determine the executable instruction set and exclude device conflict instructions through the hardware interface status detection of the in-vehicle control terminal.

[0159] The hardware interfaces of the vehicle-mounted control terminal include a braking system interface, a speed control system interface, a warning system interface, etc. Before issuing a risk control decision instruction, the status of these hardware interfaces needs to be detected. For example, it is necessary to detect whether the braking system interface is working properly. If there is a fault in the braking system, then the emergency braking instruction cannot be executed and needs to be excluded from the set of executable instructions. For the speed control system interface, if it is performing other speed adjustment operations (such as when the adaptive cruise control system is running), it is necessary to check whether there is a conflict with the speed limit control instruction. If there is a conflict, adjustment or exclusion is required. If there is a fault in the warning system interface, such as the voice warning function not working properly, then the voice reminder part in the driver warning instruction needs to be adjusted or excluded, and only the feasible parts such as the screen display reminder are retained.

[0160] Step S155, logically combine the basic risk control strategy type, the additional control parameters, the strategy effective range, and the set of executable instructions to generate the risk control decision instruction set.

[0161] For example, the finally generated risk control decision instruction set may be: within 200 meters ahead and within 10 minutes (strategy effective range), if the vehicle speed is higher than 30 km / h (speed limit control instruction, additional control parameter), and the braking system interface is normal (set of executable instructions), then execute the speed limit control instruction, and at the same time display "Low visibility ahead, please drive carefully" on the screen (driver warning instruction, additional control parameter).

[0162] In a possible implementation manner, step S153 includes:

[0163] Step S1531, obtain the topological structure data and real-time traffic flow information of the road where the target vehicle is currently located.

[0164] Step S1532, extract the road curvature radius, slope parameter, and lane line type according to the high-precision map information.

[0165] Step S1533, calculate the vehicle density distribution and average driving speed in combination with the real-time traffic flow information.

[0166] Step S1534, generate a bend control radius based on the road curvature radius and the risk level quantization index.

[0167] Step S1535, generate the priority of the avoidance area according to the vehicle density distribution and the risk event warning label.

[0168] Step S1536, integrate the bend control radius, the priority of the avoidance area, and the lane line type to define the spatial boundary and time validity period of the strategy effective range.

[0169] Assume that the target vehicle is located on the main road of the city. The road topology data shows that this is a multi-lane road with multiple intersections. The real-time traffic flow information indicates that the current traffic flow is large and the vehicle driving speed is relatively slow. Extract the road curvature radius, slope parameter, and lane line type according to the high-precision map information. If the road curvature radius is small, it indicates that there is a bend in the road, and the slope is relatively steep. The lane line type is two-way four lanes with clear lane dividers and turning signs. Combine the real-time traffic flow information to calculate the vehicle density distribution and average driving speed. In the case of large traffic flow, the vehicle density distribution is high and the average driving speed is low. Generate a bend control radius based on the road curvature radius and the risk level quantification index. If the risk level quantification index is high and the road curvature radius is small, in order to ensure the safety of the vehicle at the bend, a smaller bend control radius needs to be generated. For example, further reduce the bend control radius corresponding to the safe passing speed of the bend designed according to the road to reduce the driving speed of the vehicle at the bend and avoid the risk of skidding or losing control. Generate the priority of the avoidance area according to the vehicle density distribution and the risk event warning label. If the risk event warning label is "small vehicle spacing - medium risk" and the vehicle density distribution is high, then the priority of the avoidance area will be increased. For example, give priority to avoiding vehicles in the right lane because there may be more safety space in the right lane or it is more convenient for vehicles to perform avoidance operations. Integrate the bend control radius, the priority of the avoidance area, and the lane line type, and define the spatial boundary and time validity period of the policy effective range. For example, within the road range of the next 200 meters (spatial boundary) and within the next 10 minutes (time validity period), vehicles in these areas need to drive according to the bend control radius, and give priority to vehicle operations according to the priority of the avoidance area, and at the same time follow the regulations of the lane line type.

[0170] In one possible implementation, the method further includes:

[0171] Step S310, receive the risk control decision instruction set of multiple target vehicles through the vehicle networking cloud platform, and perform a global risk situation analysis.

[0172] Step S320, generate a regional risk heat map according to the geographical location distribution of the multiple target vehicles.

[0173] Step S330, adjust the traffic signal control strategy and the roadside unit broadcast content based on the regional risk heat map.

[0174] Step S340, send the adjusted traffic signal control strategy to the intelligent transportation management system, and update the roadside unit broadcast content.

[0175] Step S350, generate a macro vehicle networking risk trend report according to the aggregation result of the risk level quantification index of the multiple target vehicles.

[0176] For example, the vehicle networking cloud platform receives risk control decision instruction sets of multiple electric vehicles in different regions. Specifically, these vehicles are distributed in various corners of the city, some on main roads, some on branch roads, some on roads near commercial areas, and some on roads near residential areas. A regional risk heat map is generated based on the geographical location distribution of multi-target vehicles. On the roads near commercial areas, due to high vehicle density and complex traffic conditions, it may be shown as a high-risk area (red) on the regional risk heat map; while on the roads near residential areas, there are relatively fewer vehicles and slower speeds, which may be shown as low-risk areas (green). Adjust the traffic signal control strategy and the broadcast content of the roadside unit based on the regional risk heat map. For high-risk areas (such as roads near commercial areas), adjust the traffic signal control strategy, for example, extend the red light time to reduce the vehicle inflow, relieve traffic congestion and reduce the collision risk. At the same time, update the broadcast content of the roadside unit, such as broadcasting prompt information like "The road in front of the commercial area is congested, please drive carefully". Send the adjusted traffic signal control strategy to the intelligent transportation management system so that the traffic signals operate according to the adjusted strategy, for example, change the duration setting of the signal lights. And update the broadcast content of the roadside unit to timely inform the driver of relevant risk information and driving suggestions. Generate a macro vehicle networking risk trend report based on the aggregation result of the risk level quantification indicators of multi-target vehicles. For example, if the risk level quantification indicators of multiple vehicles show an upward trend over a period of time, it indicates that the risk in the entire vehicle networking area is increasing. The macro vehicle networking risk trend report can analyze that the risk increase is due to reasons such as weather changes (such as reduced visibility and slippery road surface caused by rainy weather), traffic flow peaks during specific time periods, and road construction, and put forward corresponding suggestions, such as increasing road patrols and strengthening traffic guidance measures to ensure traffic safety in the entire vehicle networking area.

[0177] Figure 2 FIG. shows a schematic diagram of exemplary hardware and software components of a vehicle networking-based risk control big data mining system 100 that can implement the ideas of the present application provided by some embodiments of the present application. For example, the processor 120 can be used on the vehicle networking-based risk control big data mining system 100 and is used to execute the functions in the present application.

[0178] The vehicle networking-based risk control big data mining system 100 can be a general-purpose server or a special-purpose server, both of which can be used to implement the vehicle networking-based risk control big data mining method of the present application. Although only one server is shown in the present application, for convenience, the functions described in the present application can be implemented in a distributed manner on multiple similar platforms to balance the processing load.

[0179] For example, the risk control big data mining system 100 based on the vehicle networking can include a network port 110 connected to a network, one or more processors 120 for executing program instructions, a communication bus 130, and different forms of storage media 140, such as disks, ROM, or RAM, or any combination thereof. Exemplarily, the risk control big data mining system 100 based on the vehicle networking can further include program instructions stored in ROM, RAM, or other types of non-transitory storage media, or any combination thereof. The method of the present application can be implemented according to these program instructions. The risk control big data mining system 100 based on the vehicle networking further includes an input / output (I / O) interface 150 between the computer and other input / output devices.

[0180] For ease of description, only one processor is described in the risk control big data mining system 100 based on the vehicle networking. However, it should be noted that the risk control big data mining system 100 in the present application can further include multiple processors. Therefore, the steps executed by one processor described in the present application can also be jointly executed or separately executed by multiple processors. For example, if the processor of the risk control big data mining system 100 based on the vehicle networking executes step A and step B, it should be understood that step A and step B can also be jointly executed by two different processors or separately executed in one processor. For example, the first processor executes step A, the second processor executes step B, or the first processor and the second processor jointly execute steps A and B.

[0181] In addition, an embodiment of the present invention further provides a readable storage medium, in which computer-executable instructions are preset. When the processor executes the computer-executable instructions, the above-mentioned risk control big data mining method based on the vehicle networking is implemented.

[0182] It should be noted that, in order to simplify the description of the present invention disclosure and thus help the understanding of one or more embodiments of the invention, in the foregoing description of the embodiments of the present invention, sometimes multiple features are merged into one embodiment, drawing, or description thereof.

Claims

1. A risk control big data mining method based on Internet of Vehicles, characterized in that: The method comprises: Collecting real-time operation data streams generated by the target vehicle in the Internet of Vehicles environment, the real-time operation data streams including vehicle status time series data, driving behavior event sets, external environment perception data, and vehicle communication interaction records; extracting a combined operation feature vector from the vehicle state time series data, extracting an abnormal behavior pattern identifier from the driving behavior event set, parsing a dynamic environmental risk factor from the external environment perception data, and converting the vehicle communication interaction record into a structured communication feature matrix; Based on the combined operation feature vector, the abnormal behavior pattern identifier, the dynamic environment risk factor and the structured communication feature matrix, construct a spatiotemporal correlation risk analysis map; Performing a dynamic driving risk assessment on the target vehicle through the spatiotemporal correlation risk analysis graph to generate a risk level quantitative index and a risk event warning label; Generate a risk control decision instruction set according to the risk level quantitative index and the risk event warning label, and feed back the risk control decision instruction set to the on-board control terminal of the target vehicle and the Internet of Vehicles cloud platform; The constructing of a spatiotemporal correlation risk analysis graph based on the combined operation feature vector, the abnormal behavior pattern identifier, the dynamic environment risk factor and the structured communication feature matrix includes: Mapping the combined operation feature vector to the state attribute node of the target vehicle, mapping the abnormal behavior pattern identifier to the driving behavior association edge, mapping the dynamic environment risk factor to the environment constraint layer, and mapping the structured communication feature matrix to the communication interaction path; Performing time slicing division on the state attribute nodes based on a preset time window to generate a time dimension topological relationship; Performing spatial grid division on the environmental constraint layer according to the vehicle geographic location data to generate a spatial dimension association matrix; Connecting the state attribute nodes in the same time slice through the driving behavior association edge to form a driving behavior influence link; Cross-layer coupling of the communication interaction path and the driving behavior impact link to generate a communication delay risk propagation channel; The time dimension topological relationship, the space dimension association matrix, the driving behavior impact link and the communication delay risk propagation channel are integrated to form the time-space association risk analysis map.

2. The risk control big data mining method based on the Internet of Vehicles according to claim 1 is characterized in that: The vehicle state time series data includes an engine speed sequence, a brake pressure waveform, a battery temperature curve, and a tire wear increment, and extracting a combined operation feature vector from the vehicle state time series data includes: Performing frequency domain transformation on the engine speed sequence to extract periodic fluctuation amplitude characteristics and speed mutation detection marks; Performing segmented integration processing on the brake pressure waveform to generate a brake intensity distribution histogram and a pressure abnormal fluctuation threshold; Performing trend decomposition on the battery temperature curve to separate long-term temperature rise gradient characteristics and short-term temperature oscillation parameters; Modeling the wear rate of the tire wear increment to calculate the tire life prediction index and the probability of abnormal tread wear; The periodic fluctuation amplitude characteristics, the speed mutation detection mark, the braking intensity distribution histogram, the pressure abnormal fluctuation threshold, the long-term temperature rise gradient characteristics, the short-term temperature oscillation parameters, the tire life prediction index and the tread abnormal wear probability are normalized and spliced ​​to form the combined operation feature vector.

3. The risk control big data mining method based on the Internet of Vehicles according to claim 1 is characterized in that: The extracting of abnormal behavior pattern identifiers from the driving behavior event set includes: The number of sudden acceleration event triggering, the time interval of continuous lane change events, the proportion of speeding duration, and the frequency of illegal light use at night are selected from the driving behavior event set; Generating a sudden acceleration risk coefficient based on the comparison of the sudden acceleration event triggering times with a preset acceleration threshold; determining a lane-changing behavior density score according to the statistical distribution of time intervals between the consecutive lane-changing events; Combining the speeding duration ratio with the road speed limit data to generate a speeding violation accumulation amount; Generate a lighting violation pattern identifier by matching the nighttime illegal lighting usage frequency with the lighting specification database; The sudden acceleration risk coefficient, the lane change behavior density score, the speeding violation accumulation amount and the light violation pattern identifier are weighted and fused to generate the abnormal behavior pattern identifier.

4. The risk control big data mining method based on the Internet of Vehicles according to claim 1 is characterized in that: The step of analyzing the dynamic environmental risk factors from the external environmental perception data includes: Identify real-time meteorological condition parameters, road obstacle distribution heat map, traffic sign semantic information and surrounding vehicle movement trajectory set from the external environment perception data; Calculating the visibility attenuation coefficient and the road adhesion risk index according to the real-time meteorological condition parameters; Generating an obstacle collision probability distribution map based on the density gradient of the road obstacle distribution heat map; Performing compliance verification on the traffic sign semantic information and generating a sign violation risk weight; Performing relative speed cluster analysis on the surrounding vehicle motion trajectory set to generate a vehicle spacing safety assessment matrix; The visibility attenuation coefficient, the road adhesion risk index, the obstacle collision probability distribution map, the sign violation risk weight and the vehicle spacing safety assessment matrix are spatially superimposed to generate the dynamic environment risk factor.

5. The risk control big data mining method based on Internet of Vehicles according to claim 1 is characterized in that: The dynamic evaluation of the driving risk of the target vehicle is performed through the spatiotemporal correlation risk analysis graph to generate a risk level quantitative index and a risk event warning label, including: Traversing the driving behavior impact links in the spatiotemporal correlation risk analysis graph, and calculating the risk transmission intensity of each link; Determining a risk response hysteresis coefficient according to the path length and delay time of the communication delay risk propagation channel; Generate an environmental risk superposition effect value based on the spatial grid density and risk factor weights of the environmental constraint layer; Performing real-time change rate monitoring on the characteristic vector of the state attribute node to generate a vehicle state degradation index; Input the risk transmission intensity, the risk response hysteresis coefficient, the environmental risk superposition effect value and the vehicle state degradation index into a risk quantification model, and output the risk level quantification index; According to the comparison between the risk level quantitative index and the preset threshold, the risk event warning labels of different levels are triggered.

6. The risk control big data mining method based on the Internet of Vehicles according to claim 5 is characterized in that: The steps of constructing the risk quantification model include: Obtain risk event sample sets annotated in historical Internet of Vehicles data, and extract the combined operation feature vectors, abnormal behavior pattern identifiers, dynamic environment risk factors, and structured communication feature matrices corresponding to the samples; Constructing an initial risk assessment neural network, wherein the initial risk assessment neural network includes a feature fusion layer, a temporal attention layer, and a spatial convolution layer; Cross-modally fusing the combined operation feature vector and the dynamic environmental risk factor through the feature fusion layer to generate an environmental coupling feature; Modeling the time dependency of the abnormal behavior pattern identifier through the time attention layer to generate a behavior timing weight; Performing spatial feature extraction on the structured communication feature matrix through the spatial convolution layer to generate a communication risk distribution map; Performing full connection processing on the environmental coupling feature, the behavior timing weight and the communication risk distribution map, and outputting a risk prediction value; Based on the back propagation of the error between the risk prediction value and the actual risk annotation value, the parameters of the initial risk assessment neural network are optimized to obtain the risk quantification model.

7. The risk control big data mining method based on the Internet of Vehicles according to claim 1 is characterized in that: The generating of the risk control decision instruction set according to the risk level quantitative index and the risk event warning label includes: Determine the type of basic risk control strategy based on the risk level quantitative indicators, including emergency braking instructions, speed limit control instructions, path re-planning instructions or driver warning instructions; In combination with the label type of the risk event warning label, additional control parameters are matched from the Internet of Vehicles strategy library; Based on the real-time location data of the target vehicle and high-precision map information, generate a strategy effective range under geographical constraints; Determine the executable instruction set and eliminate device conflicting instructions through hardware interface status detection of the vehicle control terminal; Logically combining the basic risk control strategy type, the additional control parameters, the strategy effective range, and the executable instruction set to generate the risk control decision instruction set; The strategy effectiveness range under the geographical constraint condition is generated based on the real-time location data of the target vehicle and the high-precision map information, including: Obtaining topological structure data and real-time traffic flow information of the road where the target vehicle is currently located; Extracting road curvature radius, slope parameters and lane line type according to the high-precision map information; Calculating vehicle density distribution and average driving speed in combination with the real-time traffic flow information; Generating a curve control radius based on the road curvature radius and the risk level quantitative index; generating an avoidance area priority according to the vehicle density distribution and the risk event warning label; The curve control radius, the avoidance area priority and the lane line type are integrated to define the spatial boundary and time validity period of the strategy effective range.

8. The risk control big data mining method based on the Internet of Vehicles according to claim 1 is characterized in that: The method further comprises: Receive risk control decision instruction sets of multiple target vehicles through the Internet of Vehicles cloud platform and conduct global risk situation analysis; Generating a regional risk heat map according to the geographical location distribution of the multiple target vehicles; Adjusting the traffic signal control strategy and the roadside unit broadcast content based on the regional risk heat map; Sending the adjusted traffic signal control strategy to the intelligent traffic management system and updating the roadside unit broadcast content; Based on the aggregation results of the quantitative indicators of the risk levels of the multiple target vehicles, a macro Internet of Vehicles risk trend report is generated.

9. A risk control big data mining system based on Internet of Vehicles, characterized in that: The risk control big data mining system based on the Internet of Vehicles includes a processor and a memory, the memory is connected to the processor, the memory is used to store programs, instructions or codes, and the processor is used to execute the programs, instructions or codes in the memory to implement the risk control big data mining method based on the Internet of Vehicles as described in any one of claims 1 to 8.

Images