Method, medium and device for judging dynamic risk in vehicle driving process based on big data

Abstract

The invention discloses a method, medium and device for determining dynamic risks in a vehicle driving process based on big data, and the method comprises the following steps: obtaining historical data in a vehicle driving process, carrying out the edge calculation and data preprocessing of the historical data, and obtaining target data; feature extraction is carried out based on the target data to obtain input features of the target data, and the feature extraction comprises time sequence feature extraction, environment feature extraction and behavior pattern extraction; a dynamic risk assessment model is constructed, the dynamic risk assessment model is trained according to the input features of the target data, and the dynamic risk assessment model utilizes an ARIMA model and an LSTM network model to predict dynamic risks; real-time data processing and edge calculation are provided, data processing is carried out on vehicle-mounted equipment, delay and bandwidth requirements are reduced, and the system response speed is increased. And dynamically adjusting the risk model: automatically adjusting the risk assessment model by using machine learning and deep learning technologies, and improving the prediction accuracy.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle driving risk determination, and in particular to a method, medium and device for determining dynamic risks in a vehicle driving process based on big data. Background Art

[0002] Safety is undoubtedly the top priority when a vehicle is on the road. The factors affecting safety are complex and diverse, covering drivers, vehicles, road conditions, weather and other aspects. It is worth noting that most of these factors are difficult to quantify. They are like time bombs hidden in the dark, adding many uncertain risks to the on-road process, which in turn has an unignorable impact on on-road safety. In view of this, in order to improve the safety of vehicles on the road, we need to pay attention to and analyze these factors.

[0003] There are still many technical problems in the prior art: 1. Data quality and integrity issues: Incomplete data: Data from vehicle and environmental sensors may be missing, erroneous, or inconsistent. Missing critical data or poor data quality may lead to inaccurate risk assessment results.

[0004] Data delay: There may be delays in data collection and transmission, affecting real-time performance and accuracy.

[0005] 2. Data fusion and processing problems: Data heterogeneity: Data from different sources may have inconsistent formats, scales, and qualities, making them difficult to integrate effectively.

[0006] Computational complexity: Big data analysis requires high-performance computing resources, especially in real-time environments, which places high demands on computing and storage resources.

[0007] 3. Model generalization ability: Overfitting problem: Machine learning models may overfit the training data, resulting in poor performance in practical applications.

[0008] 4. Dynamic changes: The dynamic changes of vehicles and road environments pose challenges to the adaptability of the model, which requires continuous updating and adjustment of the model. Summary of the invention

[0009] In order to solve one of the above technical problems, a method, medium and device for determining dynamic risks in vehicle driving based on big data are provided, which are used to monitor and predict potential risks in vehicle driving in real time. In order to improve the accuracy, real-time performance and user experience of existing technical solutions.

[0010] In order to achieve the above purpose, the technical solution adopted by the present invention is: A first aspect: A method for determining dynamic risks during vehicle driving based on big data, comprising the following steps: Acquire historical data of the vehicle trip, perform edge computing and data preprocessing on the historical data, and obtain target data; Performing feature extraction based on the target data to obtain input features of the target data, wherein the feature extraction includes time series feature extraction, environmental feature extraction, and behavior pattern extraction; Constructing a dynamic risk assessment model, training the dynamic risk assessment model according to the input features of the target data, and predicting dynamic risks using an ARIMA model and an LSTM network model; Generate a risk score according to the risk value output by the dynamic risk assessment model, set a threshold according to the risk score, trigger a real-time warning when the threshold is exceeded, and evaluate the safety level of the current driving state; Adjust the model warning parameters according to real-time data and risk assessment results, and optimize the driver's behavior pattern through the Q-learning algorithm.

[0011] Preferably: edge computing and data preprocessing are performed on the historical data to obtain target data, specifically including: Perform preliminary data processing on the vehicle computing platform and use edge computing nodes to perform preliminary analysis and processing of the data; Data preprocessing: clean and normalize the data, handle missing values ​​and outliers, and extract the features of acceleration and speed. The expression is: ; ; in, is the acceleration value measured at the i-th time; is the velocity value measured at the ith time.

[0012] Preferably, the input features of the target data include speed, acceleration, vehicle distance, road condition, and weather; the high-dimensional feature space is reduced in dimension by using the PCA (principal component analysis) feature extraction algorithm, and the expression is: ; in, is the data matrix, is the principal component matrix.

[0013] Preferably, the dynamic risk assessment model uses the ARIMA model and the LSTM network model to predict dynamic risks, specifically including: The ARIMA model is a random forest, and the expression is: ; in is the prediction result of the i-th decision tree for input x, is the number of decision trees; The LSTM network model is expressed as: ; in is the hidden state at the current time step, is the input data for the current time step.

[0014] Preferably, the risk score is generated according to the risk value output by the dynamic risk assessment model, a threshold is set according to the risk score, a real-time warning is triggered when the threshold is exceeded, and the safety level of the current driving state is evaluated, specifically including: The risk score uses the risk score formula to calculate the current risk, and the expression is: ; in, is the comprehensive risk score, is the weight of the feature, is the impact of feature i on risk; The real-time warning uses the threshold judgment formula: ; The safety level of the current driving state is evaluated as follows: ; in, is the weight coefficient, obtained through model training.

[0015] Preferred: The driver's behavior pattern is optimized by the Q-learning algorithm, and the expression is: ; in, is the state-action value function, It’s an instant reward. is the discount factor, is the learning rate.

[0016] Preferably: the historical data includes vehicle internal data, external environment data, big data platform data and driving behavior data, the time series feature extraction is to extract the dynamic characteristics of the vehicle; the environmental feature extraction is to extract the characteristics of environmental changes, and the behavior pattern extracts the driver's behavior pattern.

[0017] A second aspect: A device for determining dynamic risks during vehicle driving based on big data, comprising: A task acquisition module is used to acquire historical data during the vehicle journey, perform edge computing and data preprocessing on the historical data, and obtain target data; A feature extraction module, used for extracting features from the target data to obtain input features of the target data, wherein the feature extraction includes time series feature extraction, environmental feature extraction, and behavior pattern extraction; A dynamic risk assessment module is used to construct a dynamic risk assessment model, and train the dynamic risk assessment model according to the input characteristics of the target data, wherein the dynamic risk assessment model predicts dynamic risks using an ARIMA model and an LSTM network model; A risk scoring module, used to generate a risk score according to the risk value output by the dynamic risk assessment model, set a threshold according to the risk score, trigger a real-time warning when the threshold is exceeded, and evaluate the safety level of the current driving state; The driving strategy optimization module is used to adjust the model warning parameters according to real-time data and risk assessment results, and optimize the driver's behavior pattern through the Q-learning algorithm.

[0018] A third aspect: A computer device comprising: processor; A memory for storing executable instructions; Among them, the processor is used to read the executable instructions from the memory and execute the executable instructions to implement the method of determining dynamic risks during vehicle driving based on big data.

[0019] A fourth aspect: A computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the processor implements the method for determining dynamic risks during vehicle driving based on big data.

[0020] Compared with the prior art, the present invention has the following beneficial effects: The present invention proposes real-time data processing and edge computing: data processing is performed on vehicle-mounted equipment to reduce latency and bandwidth requirements and improve system response speed.

[0021] Dynamically adjust risk models: Use machine learning and deep learning technologies to automatically adjust risk assessment models to improve prediction accuracy.

[0022] Intelligent warning system: optimizes warning strategies and provides personalized driving suggestions through real-time data and user feedback.

[0023] Reinforcement learning to optimize decision-making: Use reinforcement learning algorithms to optimize driving decisions and improve the adaptability and intelligence of the system.

[0024] Through these optimizations, the solution not only improves the accuracy and real-time performance of risk assessment, but also greatly enhances the intelligence of the system and user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 A flow chart of a method for determining dynamic risks in a vehicle driving process based on big data according to the present invention; Figure 2 This is a schematic diagram of the computer device structure of Example 3 of the present invention.

[0026] Figure 3 It is a module diagram of the device for determining dynamic risks in the driving process of a vehicle based on big data according to the present invention; In the figure, 10 is a computer device; 1002 is a processor; 1004 is a memory; and 1006 is a transmission device. DETAILED DESCRIPTION

[0027] The following description is used to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments described below are only examples, and those skilled in the art may think of other obvious variations.

[0028] Embodiment 1: Reference Figure 1 As shown, a method for determining dynamic risks in a vehicle driving process based on big data includes the following steps: Acquire historical data of the vehicle trip, perform edge computing and data preprocessing on the historical data, and obtain target data; Data collection: Collect data from various sources, including vehicle sensors (such as speed, acceleration, brake status, etc.), environmental sensors (such as weather, road conditions, etc.), driving behavior data (such as sudden acceleration, sudden braking, etc.), and other traffic-related data (such as traffic signals, road conditions, etc.); Multi-source data fusion: Use data fusion algorithms to integrate data from different sources into a unified data model. Data fusion algorithms (such as Kalman filtering and particle filtering) can be used to improve the integrity and accuracy of data; Performing feature extraction based on the target data to obtain input features of the target data, wherein the feature extraction includes time series feature extraction, environmental feature extraction, and behavior pattern extraction; Time series feature extraction: Extract the dynamic characteristics of the vehicle, such as speed change rate, acceleration change rate, etc.

[0029] Environmental feature extraction: Extract features of environmental changes, such as weather trends, road condition changes, etc.

[0030] Behavioral pattern analysis: Extract driver behavior patterns, such as frequency of emergency braking, speeding behavior, etc.

[0031] Constructing a dynamic risk assessment model, training the dynamic risk assessment model according to the input features of the target data, and predicting dynamic risks using an ARIMA model and an LSTM network model; Generate a risk score according to the risk value output by the dynamic risk assessment model, set a threshold according to the risk score, trigger a real-time warning when the threshold is exceeded, and evaluate the safety level of the current driving state; Adjust the model warning parameters according to real-time data and risk assessment results, and optimize the driver's behavior pattern through the Q-learning algorithm.

[0032] Specifically: edge computing and data preprocessing are performed on the historical data to obtain target data, specifically including: Perform preliminary data processing on the vehicle computing platform and use edge computing nodes to perform preliminary analysis and processing of the data; Data preprocessing: clean and normalize the data, handle missing values ​​and outliers, and extract the features of acceleration and speed. The expression is: ; ; in, is the acceleration value measured at the i-th time; is the velocity value measured at the ith time.

[0033] Edge computing: Perform preliminary data processing on the vehicle computing platform to reduce latency and bandwidth consumption. Use edge computing nodes to perform preliminary analysis and processing of data.

[0034] Data preprocessing: cleaning, normalizing and feature extraction of data.

[0035] Specifically: the input features of the target data include speed, acceleration, vehicle distance, road conditions, and weather; the high-dimensional feature space is reduced in dimension by using the PCA (principal component analysis) feature extraction algorithm, and the expression is: ; in, is the data matrix, is the principal component matrix.

[0036] Specifically: The dynamic risk assessment model uses the ARIMA model and the LSTM network model to predict dynamic risks, specifically including: The ARIMA model is a random forest, and the expression is: ; in is the prediction result of the i-th decision tree for input x, is the number of decision trees; The LSTM network model is expressed as: ; in is the hidden state at the current time step, is the input data for the current time step.

[0037] Specifically: generating a risk score according to the risk value output by the dynamic risk assessment model, setting a threshold according to the risk score, triggering a real-time warning when the threshold is exceeded, and evaluating the safety level of the current driving state, specifically including: The risk score uses the risk score formula to calculate the current risk, and the expression is: ; in, is the comprehensive risk score, is the weight of the feature, is the impact of feature i on risk; The real-time warning uses the threshold judgment formula: ; The safety level of the current driving state is evaluated as follows: ; in, is the weight coefficient, obtained through model training.

[0038] Specifically: The driver's behavior pattern is optimized through the Q-learning algorithm, and the expression is: ; in, is the state-action value function, It’s an instant reward. is the discount factor, is the learning rate.

[0039] Real-time warning: Based on the calculated risk score, the vehicle system will issue a real-time warning to the driver. For example, It takes the form of sound prompts, visual warnings (such as HUD displays), etc.

[0040] Adaptive adjustment: Adjust the warning strategy through feedback mechanisms, such as adjusting the warning intensity under different road conditions.

[0041] Use reinforcement learning algorithms to optimize driving decisions and provide personalized driving suggestions. For example, use Q-learning algorithms to optimize driving strategies.

[0042] Specifically: the historical data includes vehicle internal data, external environment data, big data platform data and driving behavior data; the time series feature extraction is to extract the dynamic features of the vehicle; the environmental feature extraction is to extract the features of environmental changes; and the behavior pattern extracts the driver's behavior pattern.

[0043] 1. Real-time data processing and edge computing: Edge computing: Moving data processing to the on-board computing platform reduces data transmission delays and reduces reliance on cloud computing resources. This approach effectively shortens response time, enables dynamic risk assessment to be performed in real time, and improves the system’s responsiveness and efficiency.

[0044] For example: In traditional systems, data processing may be centralized in the cloud, which results in transmission delays, especially when moving at high speeds. With edge computing, vehicles can immediately process data from the CAN bus and on-board sensors, quickly identify potential risks and respond in real time, such as immediately warning the driver or adjusting driving strategies.

[0045] 2. Dynamic risk model and machine learning application: Deep learning model: Use LSTM (Long Short-Term Memory Network) to process time series data to more accurately predict dynamic risks. The LSTM model can effectively capture the long-term time dependency of vehicles during driving, improving the accuracy of risk prediction.

[0046] PCA dimensionality reduction: Principal component analysis (PCA) dimensionality reduction is used to improve the efficiency of model training and reduce the computational burden caused by high-dimensional data.

[0047] For example, LSTM can process the time series data of vehicle speed and acceleration to identify patterns that may cause risks. PCA can reduce the dimensionality of high-dimensional sensor data, thereby speeding up the training speed and prediction efficiency of the model, which is particularly important for real-time risk assessment.

[0048] 3. Intelligent early warning system and personalized configuration: Adaptive warning mechanism: Through real-time data and driver feedback, the system can dynamically adjust the warning strategy, such as automatically adjusting the warning intensity under different road conditions or weather.

[0049] Personalized configuration: Provide personalized risk management suggestions based on the driver's historical data and driving habits. This approach improves the user experience and makes the warning system more tailored to individual needs.

[0050] For example, if the system detects that the driver is slow to react in bad weather conditions, it may increase the frequency and intensity of warnings. In addition, the system can adjust the way warnings are given based on the driver's habits, such as issuing earlier warnings to drivers who often drive at high speeds.

[0051] 4. Reinforcement learning to optimize decision making Q-learning algorithm: Use reinforcement learning algorithms (such as Q-learning) to optimize driving decisions. Q-learning can continuously adjust and optimize decision-making strategies based on real-time feedback, provide personalized driving suggestions, and improve the adaptability of the system.

[0052] For example, Q-learning can optimize how to adjust the speed or acceleration strategy when encountering complex road conditions. Through continuous training and feedback, the system can provide more intelligent driving suggestions to help drivers make the best decisions in different driving environments.

[0053] Improved real-time performance and response speed: Through edge computing and real-time data processing, the system can respond to potential risks more quickly and reduce delays.

[0054] Improved accuracy of risk prediction: The use of deep learning and dimensionality reduction techniques improves the prediction accuracy of the model, making dynamic risk assessment more precise.

[0055] Optimized user experience: The intelligent early warning system and personalized configuration improve the system's adaptability and user satisfaction, making the early warning more in line with actual needs.

[0056] Enhanced the intelligence level of the system: Optimizing driving decisions through reinforcement learning, improving the system's intelligent decision-making and adaptive capabilities.

[0057] These breakthroughs enable the solution to provide more efficient and accurate risk assessment and warning in practical applications, greatly improving driving safety and user experience.

[0058] In other embodiments, the following steps are included: Data fusion algorithm: Multi-sensor data fusion: Use Kalman filter to fuse data from different sensors to improve data accuracy and reliability.

[0059] ; in, is the estimated state, is the Kalman gain, is the measured value, is the measurement matrix.

[0060] Real-time big data analysis: Stream data processing: Use tools such as Apache Kafka and Apache Flink to process real-time data streams to ensure efficient and accurate data processing.

[0061] Vehicle-to-everything (V2X) integration: V2X communication: Vehicle-to-Vehicle (V2V) and Vehicle-to-Infrastructure (V2I): Integrate V2X technology to obtain real-time information about surrounding vehicles and infrastructure to optimize dynamic risk assessment.

[0062] ; in, It is car-to-car data. It is vehicle-to-infrastructure data. It is the vehicle’s own data.

[0063] User experience optimization: Adaptive User Interface: Dynamic adjustment: Dynamically adjust the user interface display and warning mechanism based on driver feedback and driving habits. For example, use user behavior analysis to optimize the presentation of warning information.

[0064] ; in, It is a user interface adjustment. It is user feedback. It's driving mode.

[0065] Data encryption and privacy protection: Encryption technology: Use end-to-end encryption to protect data transmission security and prevent unauthorized access or tampering.

[0066] ; in, is the encrypted data. is the encryption function, is the encryption key, is the original data.

[0067] Further improve the overall performance, user experience and security of the system.

[0068] Embodiment 2 Figure 3 As shown: A device for determining dynamic risks during vehicle driving based on big data, comprising: A task acquisition module is used to acquire historical data during the vehicle journey, perform edge computing and data preprocessing on the historical data, and obtain target data; A feature extraction module, used for extracting features from the target data to obtain input features of the target data, wherein the feature extraction includes time series feature extraction, environmental feature extraction, and behavior pattern extraction; A dynamic risk assessment module is used to construct a dynamic risk assessment model, and train the dynamic risk assessment model according to the input characteristics of the target data, wherein the dynamic risk assessment model predicts dynamic risks using an ARIMA model and an LSTM network model; A risk scoring module, used to generate a risk score according to the risk value output by the dynamic risk assessment model, set a threshold according to the risk score, trigger a real-time warning when the threshold is exceeded, and evaluate the safety level of the current driving state; The driving strategy optimization module is used to adjust the model warning parameters according to real-time data and risk assessment results, and optimize the driver's behavior pattern through the Q-learning algorithm.

[0069] The foregoing Figure 1 The various variations and specific examples of the method for determining dynamic risks in a vehicle driving process based on big data in Example 1 are also applicable to a device for determining dynamic risks in a vehicle driving process based on big data in this embodiment. Through the above detailed description of the method for determining dynamic risks in a vehicle driving process based on big data, those skilled in the art can clearly know the implementation method of the device for determining dynamic risks in a vehicle driving process based on big data in this embodiment. Therefore, for the sake of brevity of the specification, it will not be described in detail here.

[0070] Example 3 An embodiment of the present application provides a computer device, which includes a processor and a memory, wherein the memory stores at least one instruction or at least one program, and the at least one instruction or the at least one program is loaded and executed by the processor to implement a method for determining dynamic risks during vehicle driving based on big data as provided in the above-mentioned method embodiment.

[0071] Figure 2 A schematic diagram of the hardware structure of a device for implementing a method for determining dynamic risks in a vehicle driving process based on big data provided in an embodiment of the present application is shown. The device may participate in or include the apparatus or system provided in an embodiment of the present application. Figure 2As shown, the computer device 10 may include one or more processors 1002 (the processor may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA), a memory 1004 for storing data, and a transmission device 1006 for communication functions. In addition, it may also include: a display, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of the I / O interface), a network interface, a power supply and / or a camera. It can be understood by those skilled in the art that Figure 2 The structure shown is only for illustration and does not limit the structure of the above electronic device. Figure 2 More or fewer components as shown, or with Figure 2 Different configurations shown.

[0072] It should be noted that the one or more processors and / or other data processing circuits described above may generally be referred to herein as "data processing circuits". The data processing circuits may be embodied in whole or in part as software, hardware, firmware, or any other combination thereof. In addition, the data processing circuit may be a single independent processing module, or may be incorporated in whole or in part into any of the other components in the computer device 10 (or mobile device). As described in the embodiments of the present application, the data processing circuit acts as a processor control (e.g., selection of a variable resistor terminal path connected to an interface).

[0073] The memory 1004 can be used to store software programs and modules of application software, such as a program instruction / data storage device corresponding to a method for determining dynamic risks in a vehicle driving process based on big data in an embodiment of the present application. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory 1004, that is, implementing the above-mentioned method. The memory 1004 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 1004 may further include a memory remotely arranged relative to the processor, and these remote memories may be connected to the computer device 10 via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0074] The transmission device 1006 is used to receive or send data via a network. The specific example of the above network may include a wireless network provided by a communication provider of the computer device 10. In one example, the transmission device 1006 includes a network adapter (Network Interface Controller, NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 1006 can be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0075] The display may be, for example, a touch screen liquid crystal display (LCD) that enables a user to interact with a user interface of the computer device 10 (or mobile device).

[0076] Example 4 An embodiment of the present application also provides a computer-readable storage medium, which can be set in a server to store at least one instruction or at least one program related to a method for determining dynamic risks in a vehicle driving process based on big data in a method embodiment. The at least one instruction or the at least one program is loaded and executed by the processor to implement a method for determining dynamic risks in a vehicle driving process based on big data provided in the above method embodiment.

[0077] Optionally, in this embodiment, the storage medium may be located in at least one of the multiple network servers of the computer network. Optionally, in this embodiment, the storage medium may include, but is not limited to, various media that can store program codes, such as a USB flash drive, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk, or an optical disk.

[0078] Example 5 The embodiment of the present invention also provides a computer program product or a computer program, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes a method for determining dynamic risks in a vehicle driving process based on big data provided in the above various optional embodiments.

[0079] It should be noted that the above-mentioned sequence of the embodiments of the present application is for description only and does not represent the advantages and disadvantages of the embodiments. The above-mentioned specific embodiments of the present application are described. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0080] Each embodiment in this application is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device, equipment and storage medium embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0081] A person skilled in the art will understand that all or part of the steps to implement the above embodiments may be accomplished by hardware or by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, and the above-mentioned storage medium may be a read-only memory, a disk or an optical disk, etc.

[0082] Based on the above ideal embodiments of the present invention, the relevant staff can make various changes and modifications without departing from the technical concept of the present invention through the above description. The technical scope of the present invention is not limited to the contents of the specification, and its technical scope must be determined according to the scope of the claims.

[0083] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of various changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A method for determining dynamic risks during vehicle driving based on big data, characterized in that: The following steps are involved: Acquire historical data of the vehicle trip, perform edge computing and data preprocessing on the historical data, and obtain target data; Performing feature extraction based on the target data to obtain input features of the target data, wherein the feature extraction includes time series feature extraction, environmental feature extraction, and behavior pattern extraction; Constructing a dynamic risk assessment model, training the dynamic risk assessment model according to the input features of the target data, and predicting dynamic risks using an ARIMA model and an LSTM network model; Generate a risk score according to the risk value output by the dynamic risk assessment model, set a threshold according to the risk score, trigger a real-time warning when the threshold is exceeded, and evaluate the safety level of the current driving state; Adjust the model warning parameters according to real-time data and risk assessment results, and optimize the driver's behavior pattern through the Q-learning algorithm.

2. The method according to claim 1, characterized in that: Edge computing and data preprocessing are performed on the historical data to obtain target data, specifically including: Perform preliminary data processing on the vehicle computing platform and use edge computing nodes to perform preliminary analysis and processing of the data; Data preprocessing: clean and normalize the data, process missing values ​​and outliers, and extract the features of acceleration and velocity. The expression is: ； ； in, is the acceleration value measured at the i-th time; is the velocity value measured at the ith time.

3. The method according to claim 1, characterized in that: The input features of the target data include speed, acceleration, vehicle distance, road condition, and weather. The high-dimensional feature space is reduced in dimension by the PCA (principal component analysis) feature extraction algorithm, and the expression is: ； in, is the data matrix, is the principal component matrix.

4. The method according to claim 1, characterized in that: The dynamic risk assessment model uses the ARIMA model and the LSTM network model to predict dynamic risks, specifically including: The ARIMA model is a random forest, and the expression is: ； in is the prediction result of the i-th decision tree for input x, is the number of decision trees; The LSTM network model is expressed as: ； in is the hidden state at the current time step, is the input data for the current time step.

5. The method according to claim 1, characterized in that: The risk score is generated according to the risk value output by the dynamic risk assessment model, a threshold is set according to the risk score, a real-time warning is triggered when the threshold is exceeded, and the safety level of the current driving state is evaluated, specifically including: The risk score uses the risk score formula to calculate the current risk, and the expression is: ； in, is the comprehensive risk score, is the weight of the feature, is the impact of feature i on risk; The real-time warning uses the threshold judgment formula: ； The safety level of the current driving state is evaluated as follows: ； in, is the weight coefficient, obtained through model training.

6. The method according to claim 1, characterized in that: The driver's behavior pattern is optimized by the Q-learning algorithm, and the expression is: ； in, is the state-action value function, It’s an instant reward. is the discount factor, is the learning rate.

7. The method according to claim 1, characterized in that: The historical data includes vehicle internal data, external environment data, big data platform data and driving behavior data. The time series feature extraction is to extract the dynamic features of the vehicle; the environmental feature extraction is to extract the features of environmental changes; and the behavior pattern extracts the pattern of driver behavior.

8. A device for determining dynamic risks during vehicle driving based on big data, characterized in that: include: A task acquisition module is used to acquire historical data during the vehicle journey, perform edge computing and data preprocessing on the historical data, and obtain target data; A feature extraction module, used for extracting features from the target data to obtain input features of the target data, wherein the feature extraction includes time series feature extraction, environmental feature extraction, and behavior pattern extraction; A dynamic risk assessment module is used to construct a dynamic risk assessment model, and train the dynamic risk assessment model according to the input characteristics of the target data, wherein the dynamic risk assessment model predicts dynamic risks using an ARIMA model and an LSTM network model; A risk scoring module, used to generate a risk score according to the risk value output by the dynamic risk assessment model, set a threshold according to the risk score, trigger a real-time warning when the threshold is exceeded, and evaluate the safety level of the current driving state; The driving strategy optimization module is used to adjust the model warning parameters according to real-time data and risk assessment results, and optimize the driver's behavior pattern through the Q-learning algorithm.

9. A computer device, characterized in that: include: processor; A memory for storing executable instructions; Wherein, the processor is used to read the executable instructions from the memory and execute the executable instructions to implement the method for determining dynamic risks during vehicle driving based on big data as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the processor implements the method for determining dynamic risks in a vehicle driving process based on big data as described in any one of claims 1 to 7.

Images