Search

CN-121526003-B - Emergency rescue linkage method and system for analyzing vehicle state based on driving behavior

CN121526003BCN 121526003 BCN121526003 BCN 121526003BCN-121526003-B

Abstract

The invention discloses an emergency rescue linkage method and system for analyzing a vehicle state based on driving behaviors, and relates to the field of big data processing, wherein the emergency rescue linkage method comprises the steps of collecting driving behaviors, vehicle operation and environment data in real time, constructing a standardized data set, training a normal driving behavior reference model by using a machine learning algorithm, and setting a characteristic threshold range; after the real-time data is compared with the model, the abnormal grade is judged through the deviation value, the abnormal source is analyzed by combining the vehicle operation and the environment data, the safety risk is evaluated, whether emergency rescue is needed is judged, if rescue is needed, the integrated data are uploaded to a cloud platform, the platform links a rescue mechanism, medical treatment, traffic management and insurance companies according to the vehicle position, the abnormal grade and the peripheral resource distribution, and in the rescue process, the platform tracks the rescue progress in real time. The invention has the advantages that the accurate recognition and the root judgment of driving abnormality are realized by integrating the multidimensional data, the response efficiency and the accuracy of emergency rescue are greatly improved, and the driving safety is ensured.

Inventors

  • DENG QINGLONG

Assignees

  • 福州吉诺网络科技有限公司

Dates

Publication Date
20260508
Application Date
20260115

Claims (9)

  1. 1. The emergency rescue linkage method for analyzing the state of the vehicle based on the driving behavior is characterized by comprising the following steps: acquiring driving behavior data, vehicle operation basic data and environment data in real time based on sensor equipment on a vehicle, and acquiring a standardized data set after cleaning, denoising and standardized processing; Based on the obtained standardized driving behavior data set, extracting static features and dynamic features of driving behaviors, based on normal driving behavior sample data, training and constructing a normal driving behavior reference model by adopting a machine learning algorithm, and obtaining a feature threshold range of the normal driving behaviors through iterative training; inputting the driving behavior characteristics extracted in real time into a normal driving behavior reference model, calculating the deviation value of the real-time driving behavior characteristics and the normal driving behavior characteristics in the reference model, judging the driving abnormal behavior by setting a deviation threshold value, and marking the abnormal state grade; Determining the root type of the abnormal state of the vehicle through association analysis based on the driving behavior abnormal data, the vehicle operation abnormal data and the environment data, evaluating the safety risk value of the vehicle by adopting a risk evaluation matrix based on the root type and the abnormal state level, and carrying out emergency rescue linkage judgment; based on the linkage of emergency rescue, integrating the related data of the abnormal state of the vehicle, transmitting the data to a cloud platform through vehicle-mounted terminal equipment, and checking information; The cloud platform sends linkage data to a rescue main body, a medical institution, a traffic management department and an insurance company through linkage interfaces based on the current position of the vehicle, the abnormal state level and surrounding rescue resource distribution; In the emergency rescue implementation process, the cloud platform receives positioning information of a rescue vehicle, feedback information of rescue workers and vehicle state update data uploaded by the vehicle-mounted terminal equipment in real time, and tracks the rescue progress in real time.
  2. 2. The emergency rescue linkage method based on the analysis of the vehicle state by the driving behavior according to claim 1, wherein the acquisition of the standardized data set by the sensor device on the vehicle to acquire the driving behavior data, the vehicle operation basic data and the environmental data in real time after the cleaning, the denoising and the standardization process specifically comprises the following steps: The driving behavior data comprise steering operation data of a driver, accelerator pedal travel data, brake pedal trigger frequency and force data, gear switching data, steering wheel rotation angle and angular speed data; The vehicle operation basic data comprise engine speed data, vehicle running speed data, brake system pressure data, tire pressure data, engine oil pressure data, cooling liquid temperature data and battery voltage data; the environment data comprise current running position data of the vehicle, road condition grade data of a running road section, real-time weather data and illumination intensity data; and cleaning, denoising and standardizing the acquired data to acquire a standardized driving behavior data set, a standardized vehicle running data set and a standardized environment data set.
  3. 3. The emergency rescue linkage method based on the driving behavior analysis vehicle state according to claim 1, wherein the step of extracting static features and dynamic features of the driving behavior based on the acquired standardized driving behavior data set, and based on normal driving behavior sample data, training and constructing a normal driving behavior reference model by adopting a machine learning algorithm, and acquiring a feature threshold range of the normal driving behavior through iterative training specifically comprises: setting a time window during static feature extraction, calculating an average steering angle, an average accelerator pedal stroke and an average brake frequency in the window, and recording gear switching times in unit mileage; Calculating the instantaneous change rate of steering angle, the lifting rate of the travel of an accelerator pedal and the difference value between the peak value and the valley value of braking force in real time during dynamic feature extraction, and counting the continuous driving duration and the fluctuation times of driving behavior features in unit time to form a feature set; based on a normal driving behavior sample of a vehicle of the same model, a normal driving behavior reference model is constructed by adopting a gradient lifting tree algorithm, and the extracted static and dynamic characteristics are used as input vectors to carry out multi-round iterative training optimization to obtain a normal threshold range of each driving behavior characteristic.
  4. 4. The emergency rescue linkage method based on the analysis of the driving behavior of the vehicle according to claim 1, wherein the inputting the driving behavior feature extracted in real time into the normal driving behavior reference model, calculating the deviation value of the real-time driving behavior feature and the normal driving behavior feature in the reference model, judging the driving abnormal behavior by setting the deviation threshold value, and marking the abnormal state level specifically comprises: Inputting the static and dynamic characteristics of the driving behavior extracted in real time into a normal driving behavior reference model, calculating the deviation value of the real-time characteristics and normal characteristic reference values in the model dimension by dimension, and generating a deviation data sequence corresponding to each characteristic; Aiming at deviation data sequences of different types of characteristics, setting a differential deviation threshold value by combining a vehicle model and a driving scene, and triggering an abnormal alarm when the threshold value is exceeded; And combining the vehicle operation data set, performing secondary verification on the primarily determined driving behavior abnormality, and marking the abnormal state level of the vehicle in the abnormal state.
  5. 5. The emergency rescue linkage method based on the driving behavior analysis vehicle state according to claim 1, wherein the determining the root type of the abnormal vehicle state by the association analysis based on the driving behavior abnormal data, the vehicle operation abnormal data and the environmental data, and evaluating the safety risk value of the vehicle by using a risk evaluation matrix based on the root type and the abnormal state level, and performing the emergency rescue linkage judgment specifically comprises: based on driving behavior abnormal data, vehicle operation abnormal data and environment data, aligning the three types of data according to an abnormal occurrence time sequence, screening data dimensions strongly related to abnormal states through feature matching, and judging the root type; the root cause type includes operator mishandling, vehicle component failure, and external environmental disturbances; and constructing a risk assessment matrix based on the root type, the abnormal state level and the current running environment, quantifying risk weights of all dimensions, comprehensively calculating a safety risk value, and triggering an emergency rescue linkage mechanism when the risk value exceeds a preset threshold value.
  6. 6. The emergency rescue linkage method based on the driving behavior analysis vehicle state according to claim 1, wherein the integration of the vehicle abnormal state related data based on the emergency rescue linkage is performed, the vehicle abnormal state related data is transmitted to a cloud platform through a vehicle-mounted terminal device, and the information verification specifically comprises: When the emergency rescue linkage mechanism is judged to be started, integrating the related data of the abnormal state of the vehicle, wherein the related data comprises vehicle unique identification information, real-time positioning information, abnormal state level, abnormal root position result, vehicle operation key data, driver basic information and real-time environment data; The integrated rescue information is subjected to full-scale encryption by adopting a symmetric encryption algorithm, the encrypted information is transmitted to a cloud platform, the cloud platform analyzes the encrypted information after receiving the encrypted information, and the integrity and the accuracy of the data are verified through a check code; If the missing field or the data error exists, a directional supplementary transmission request is sent to the vehicle-mounted terminal, and after the supplementary transmission is completed and checked, a receipt confirmation receipt is generated.
  7. 7. The emergency rescue linkage method based on the driving behavior analysis vehicle state according to claim 1, wherein the cloud platform transmitting linkage data to a rescue subject, a medical institution, a traffic management department and an insurance company through a linkage interface based on a current position of the vehicle, an abnormal state level and surrounding rescue resource distribution specifically comprises: sending rescue instructions to a rescue main body through a linkage interface of the cloud platform, wherein the rescue main body comprises a nearby road rescue mechanism, a medical mechanism, a traffic management department and an insurance company; Transmitting information of vehicle positioning, abnormal states and fault components to a road rescue mechanism, and dispatching the rescue vehicle to the scene; Sending early warning information of a driver and vehicle positioning to a medical institution; Transmitting abnormal information of the vehicle and information of a driving road section to a traffic management department for traffic dispersion; sending accident early warning information to an insurance company; and sending rescue progress prompt information to the driver through the vehicle-mounted terminal equipment.
  8. 8. The emergency rescue linkage method based on the driving behavior analysis vehicle state according to claim 1, wherein in the emergency rescue implementation process, the cloud platform receives positioning information of a rescue vehicle, feedback information of a rescue person and vehicle state update data uploaded by the vehicle-mounted terminal device in real time, and the real-time tracking of the rescue progress specifically comprises: In the emergency rescue implementation process, the cloud platform receives positioning information of a rescue vehicle, feedback information of rescue personnel and vehicle state update data in real time, and tracks the rescue progress in real time; If the rescue route is jammed, the abnormal state of the vehicle is aggravated or the injury condition of the driver is changed, the cloud platform adjusts the rescue scheme and reschedules the corresponding rescue main body; When the rescue vehicle arrives at the scene and rescue treatment is completed, the cloud platform records the whole rescue process data, generates rescue summary, and simultaneously sends rescue completion notification to each linkage rescue main body to terminate the emergency rescue linkage flow.
  9. 9. Emergency rescue linkage system for analysing a vehicle condition based on driving behaviour, for implementing an emergency rescue linkage method for analysing a vehicle condition based on driving behaviour according to any one of claims 1 to 8, comprising: The data acquisition and preprocessing module is used for acquiring driving behavior, vehicle running and environment data in real time through the vehicle-mounted sensor, and performing cleaning, denoising and standardization processing to generate a structured data set; The feature extraction and model training module is used for extracting static and dynamic features of driving behaviors, constructing a normal driving behavior reference model by using a machine learning algorithm and determining a feature threshold range; the abnormal detection and grade marking module is used for calculating the deviation value of the driving behavior characteristic and the reference model in real time, judging the abnormal behavior by combining the threshold value, and marking the abnormal state grade; The root cause analysis and risk assessment module is used for associating driving behaviors, vehicle operation and environment data, identifying abnormal root cause types and judging whether emergency rescue is triggered or not through a risk assessment matrix; The data integration and transmission module is used for encrypting and integrating abnormal data of the vehicle and transmitting the abnormal data to the cloud platform to check the integrity of the data; the rescue scheduling and linkage module is used for sending linkage instructions to rescue institutions, medical institutions, traffic management departments and insurance companies according to the positions of vehicles, abnormal grades and rescue resource distribution, and coordinating rescue actions; The rescue tracking and feedback module is used for monitoring the rescue progress in real time, dynamically adjusting the rescue scheme, recording the rescue result and terminating the linkage flow; the processor is used for processing the calculation process of each formula and the construction calculation process of each model.

Description

Emergency rescue linkage method and system for analyzing vehicle state based on driving behavior Technical Field The invention relates to the field of big data processing, in particular to an emergency rescue linkage method and system for analyzing a vehicle state based on driving behaviors. Background In recent years, the vehicle CAN bus data and the driver behavior data are collected in real time, abnormal driving behaviors or vehicle faults CAN be dynamically identified by combining a machine learning algorithm, automatic early warning is triggered and linked with a rescue center, and at present, some high-end vehicle types are equipped with the system, but challenges such as insufficient data fusion precision, high false alarm rate, low multi-source information cooperative efficiency and the like generally exist. In the future, with the maturity of 5G, edge calculation and V2X technologies, the real-time performance and reliability of driving behavior analysis and emergency rescue are further improved, and the reduction of the death rate of traffic accidents and the intellectualization of emergency rescue systems are promoted. The related method for vehicle emergency rescue in the market has obvious disadvantages, and is mainly characterized by insufficient multi-dimensional cooperation and intelligent level. The system is dependent on single vehicle operation data or manual alarm to trigger rescue, lacks deep mining and integrated analysis on dynamic and static characteristics of driving behaviors, is difficult to accurately identify abnormal sources, and is easy to misjudge or miss judge. Meanwhile, information islands exist among departments, unified cloud platform scheduling is lacked, and rescue response is delayed, and cooperation of multiple main bodies is poor. The positioning aspect is easily influenced by complex terrains to generate deviation, and the rescue process lacks a dynamic tracking and scheme optimizing mechanism. In addition, the system has poor suitability for special scenes, insufficient professional capability for coping with special situations such as new energy vehicle faults and the like, and is difficult to form a complete process accurate control from abnormal early warning to rescue closed loop, and rescue efficiency and safety are limited. Disclosure of Invention In order to perfect the existing method and system, an emergency rescue linkage method and system for analyzing the state of a vehicle based on driving behaviors are provided, the method realizes accurate identification and root judgment of driving anomalies by integrating multidimensional data, and a cloud platform is relied on to link multiple departments to cooperatively rescue and dynamically optimize a scheme, so that a full-flow closed-loop management is formed, the response efficiency and the accuracy of emergency rescue are greatly improved, and the driving safety is ensured. In order to achieve the above purpose, the invention adopts the following technical scheme: an emergency rescue linkage method for analyzing a vehicle state based on driving behavior, comprising: acquiring driving behavior data, vehicle operation basic data and environment data in real time based on sensor equipment on a vehicle, and acquiring a standardized data set after cleaning, denoising and standardized processing; Based on the obtained standardized driving behavior data set, extracting static features and dynamic features of driving behaviors, based on normal driving behavior sample data, training and constructing a normal driving behavior reference model by adopting a machine learning algorithm, and obtaining a feature threshold range of the normal driving behaviors through iterative training; inputting the driving behavior characteristics extracted in real time into a normal driving behavior reference model, calculating the deviation value of the real-time driving behavior characteristics and the normal driving behavior characteristics in the reference model, judging the driving abnormal behavior by setting a deviation threshold value, and marking the abnormal state grade; Determining the root type of the abnormal state of the vehicle through association analysis based on the driving behavior abnormal data, the vehicle operation abnormal data and the environment data, evaluating the safety risk value of the vehicle by adopting a risk evaluation matrix based on the root type and the abnormal state level, and carrying out emergency rescue linkage judgment; based on the linkage of emergency rescue, integrating the related data of the abnormal state of the vehicle, transmitting the data to a cloud platform through vehicle-mounted terminal equipment, and checking information; The cloud platform sends linkage data to a rescue main body, a medical institution, a traffic management department and an insurance company through linkage interfaces based on the current position of the vehicle, the abnormal state level and surrounding resc