Driving assistance method, system and equipment and readable storage medium

By identifying and analyzing the behavior of the target driver, using the multi-agent system and Q-learning algorithm to generate the best action path, the problem of intelligent driving assistance cannot be customized is solved, and the driving behavior specifications and safety improvements are achieved.

CN119037453AActive Publication Date: 2024-11-29MOUDI INTELLIGENT TECH CO LTD

Patent Information

Application Number
CN202411446516.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-10-16
Publication Date
2024-11-29
Estimated Expiration
2044-10-16

AI Technical Summary

Technical Problem

Intelligent driving assistance technology fails to identify specific target drivers and cannot understand their driving habits, resulting in the inability to customize intelligent driving assistance for target drivers, affecting the specifications and optimization of driving behavior.

Method used

By detecting and identifying surrounding environment information and target driver behavior information, a multi-agent system is established, and driving training is used to generate optimal action paths, and these paths are executed in the current driving state to achieve personality Intelligent driving assistance.

Benefits of technology

Personalized assisted driving for target drivers is achieved, standardized and optimized driving behavior, reduced energy consumption, and improved driving safety.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119037453A_ABST
    Figure CN119037453A_ABST
Patent Text Reader

Abstract

The invention relates to a driving assistance method, which comprises the following steps of: detecting and identifying surrounding environment information of a preset vehicle and behavior information of a target driver, analyzing and processing, and extracting driving data influencing driving operation; based on an existing auxiliary driving function, simulating a preset vehicle driving process, and establishing an initial multi-agent system; and performing driving training based on the target driver on the initial multi-agent system based on an improved Q-learning algorithm to obtain an optimal action path, updating the optimal action path to the multi-agent system, directly transmitting the optimal action path to each agent, and executing the driving operation of the optimal action path according to the current driving state. The intelligent driving assistance method is customized for the target driver, and the purposes of standardizing and optimizing the driving behavior of the target driver, reducing the driving energy consumption and improving the driving safety are achieved by recognizing the specific target driver, knowing the driving habit of the target driver for a long time and strengthening the intelligent driving capacity.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent automobile assisted driving, and in particular to a driving assistance method. Background Art

[0002] A new round of technological revolution and industrial transformation is in the ascendant, and intelligent driving has become the strategic direction of the development of the global automotive industry. At present, intelligent driving has entered a new era, and intelligent driving assistance such as adaptive cruise control, lane deviation, and fatigue detection are also becoming popular. However, intelligent driving assistance currently fails to identify specific target drivers, understand the driving habits of target drivers, and customize intelligent driving assistance for target drivers. Summary of the invention

[0003] The main purpose of the present invention is to provide a driving assistance method to customize an intelligent driving assistance method for a target driver. By identifying a specific target driver and understanding his or her driving habits over a long period of time, the ability of intelligent driving can be enhanced to standardize and optimize the driving behavior of the target driver, reduce driving energy consumption, and improve driving safety.

[0004] To achieve the above object, the present invention provides a driving assistance method, comprising the following steps:

[0005] Detect and identify the surrounding environment information of the preset vehicle and the behavior information of the target driver, analyze and process the surrounding environment information and the behavior information of the target driver, and extract driving data that affects the driving operation;

[0006] Based on the existing assisted driving functions, simulate the preset vehicle driving process and establish the initial multi-agent system;

[0007] Based on the improved Q-learning algorithm, the initial multi-agent system is trained based on the target driver to obtain the optimal action path and update it to the multi-agent system;

[0008] The optimal action path is directly transmitted to each intelligent agent, and the driving operation of the optimal action path is executed according to the current driving state.

[0009] Furthermore, the steps of detecting and identifying the surrounding environment information of a preset vehicle and the target driver's behavior information, analyzing and processing the surrounding environment information and the target driver's behavior information, and extracting driving data that affects the driving operation include:

[0010] The basic data of the preset vehicle is loaded through the vehicle control system, the peripheral environment information of the preset vehicle is collected automatically through the external Tbox and radar of the preset vehicle, and the behavior data of the target driver is collected through the built-in camera and voice equipment of the preset vehicle; the basic data at least includes the license plate number, vehicle model, driver's license, the location of the preset vehicle, and the driving destination, and the surrounding environment information at least includes the driving environment data of roads, vehicles, pedestrians, and traffic signals; the built-in camera and voice equipment serve as the communication medium between the target driver and the vehicle control system;

[0011] The collected driving data is entered into the system background using data management software; the driving data includes the basic data, the driving environment data and the behavior data.

[0012] Furthermore, the step of simulating a preset vehicle driving process based on the existing assisted driving function and establishing an initial multi-agent system includes:

[0013] Based on the existing assisted driving function, simulate the driving process of a preset vehicle and perform a driving test on the driving process;

[0014] Collecting the test surrounding environment information in real time during the driving test through the sensors of the preset vehicle;

[0015] Based on the test surrounding environment information, identifying a test driving scenario;

[0016] According to the driving scenario, the system simulates the behavior patterns of each agent;

[0017] Conducting driving tests on the behavior patterns of each intelligent agent in each of the driving scenarios through simulation experiments or actual roads;

[0018] During the driving test, the behavior indicators of each intelligent agent are recorded to evaluate the functional effectiveness and safety, and an evaluation result is obtained;

[0019] An initial multi-intelligent system is established based on the evaluation results, including a driving agent, an automatic planning agent, a perception agent and a core control agent.

[0020] The step of collecting the test surrounding environment information in real time during the driving test by using the sensors of the preset vehicle includes:

[0021] Collecting the surrounding environment data of the driving test process in real time through the sensors of the preset vehicle;

[0022] The surrounding environment data is automatically screened by factors through Rule-based RewardModel to obtain the final test surrounding environment information, wherein the factor automatic screening rules at least include screening out information threatening driving safety and information related to traffic rules through the existing auxiliary driving system; the factor automatic screening rules are automatically adjusted according to the changes in the driving environment and the update of sensor data according to new environmental conditions and task requirements.

[0023] Furthermore, the step of performing driving training based on the target driver on the initial multi-agent system based on the improved Q-learning algorithm to obtain the optimal action path and update it to the multi-agent system includes:

[0024] Establish behavioral state factors and list them as Q-tables for agent operations;

[0025] Initialize the Q-table, initialize the Q value to 0, and perform driving training based on the improved Q-learning algorithm, the formula of which is:

[0026]

[0027] Among them, γ (reward), s (state), a (action), among which σ 2 is the influence of the factor of the selected agent, the initial value is 1, and the square exploration method can be used to adjust the optimal factor. It is the adjustment of the core control body;

[0028] The preset vehicle enters the test process, and according to the current vehicle state s of the preset vehicle, a driving action a1 of a preset agent is selected and operated, and then the preset agent selects driving action a2 to operate, and obtains feedback results after a1 and a2 operations; wherein the parameter of the agent parameter a1 is σ 2 ;

[0029] Compare the feedback result with the target driver behavior, the best behavior s' and the reward γ, and update it to the following formula Q(S, a);

[0030] Q(S,a)=Q(S,a)+α[R(S,a)+γmaxQ′(s′,a′)-Q(S,a)

[0031] The test process is executed one by one for each agent in each scene until the data in the Q-table is updated, the optimal action path in each driving state is obtained, and the optimal action path is updated to the multi-agent system.

[0032] Furthermore, the step of directly transmitting the optimal action path to each agent and executing the driving operation of the optimal action path according to the current driving state includes:

[0033] Determine whether to enter the driving mode;

[0034] If the driving mode is entered, the perception agent is triggered to obtain the current driving state factor;

[0035] Based on the current driving state factor, obtaining an optimal driving operation path in the current state according to the multi-agent system;

[0036] The optimal driving operation behavior instruction is transmitted to each intelligent agent through the core intelligent agent;

[0037] Determining whether to execute the optimal driving operation path based on the current behavior state of the target driver;

[0038] If the optimal driving operation path is executed, the driving operation is performed according to the procedure of the current optimal driving operation path.

[0039] The present invention also provides a driving assistance method system, comprising:

[0040] A data processing module, used to detect and identify the surrounding environment information of a preset vehicle and the behavior information of a target driver, analyze and process the surrounding environment information and the behavior information of the target driver, and extract driving data that affects driving operations;

[0041] The system establishment module is used to simulate the preset vehicle driving process and establish the initial multi-agent system based on the existing assisted driving function;

[0042] A system training module, which performs driving training based on the target driver on the initial multi-agent system according to an improved Q-learning algorithm, obtains an optimal action path and updates it to the multi-agent system;

[0043] The operation execution module directly transmits the optimal action path to each intelligent agent and executes the driving operation of the optimal action path according to the current driving state.

[0044] The present invention also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above-mentioned driving assistance method when executing the computer program.

[0045] The present invention also provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps of the above-mentioned driving assistance method are implemented.

[0046] The driving assistance method provided by the present invention has the following beneficial effects:

[0047] (1) Compared with traditional algorithms, the present invention can identify specific target drivers, understand and analyze their driving habits over a long period of time, and thus customize intelligent driving assistance methods that are more in line with individual characteristics. While providing personalized assisted driving, it also helps to standardize and optimize the driving behavior of the target driver;

[0048] (2) By extracting driving data that affects driving operations and performing intelligent analysis and processing, the present invention can guide drivers to adopt a more energy-efficient driving method to reduce energy consumption. At the same time, by simulating and training multi-agent systems, the present invention can predict and avoid potential dangerous situations, thereby improving driving safety.

[0049] (3) The present invention uses an improved Q-learning algorithm to train the multi-agent system, obtain the optimal action path, and automatically update it to the system, so that the system can automatically determine whether to execute the optimal driving operation path according to the current driving status, further improving the level of intelligence and automation. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 is a schematic flow chart of a driving assistance method in one embodiment of the present invention;

[0051] Figure 2 is a structural block diagram of a driving assistance method system in one embodiment of the present invention;

[0052] Figure 3 is a schematic diagram of a Q-table for calculating expected returns in one embodiment of the present invention;

[0053] Figure 4 It is a schematic block diagram of the structure of a computer device according to an embodiment of the present invention.

[0054] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0055] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0056] Reference Figure 1 , is a flow chart of a driving assistance method proposed by the present invention, comprising the following steps:

[0057] S1, detecting and identifying the surrounding environment information of a preset vehicle and the behavior information of a target driver, analyzing and processing the surrounding environment information and the behavior information of the target driver, and extracting driving data that affects the driving operation;

[0058] S2, based on the existing assisted driving function, simulates the preset vehicle driving process and establishes the initial multi-agent system;

[0059] S3, performing driving training based on the target driver on the initial multi-agent system based on an improved Q-learning algorithm, obtaining an optimal action path and updating it to the multi-agent system;

[0060] S4, directly transmitting the optimal action path to each intelligent agent, and executing the driving operation of the optimal action path according to the current driving state.

[0061] For S1, specifically:

[0062] S11, loading the basic data of the preset vehicle through the vehicle control system, collecting the surrounding environment information of the preset vehicle by the external Tbox and radar of the preset vehicle, and collecting the behavior data of the target driver through the built-in camera and voice equipment of the preset vehicle; the basic data at least includes the license plate number, vehicle model, driver's license, the location of the preset vehicle, and the driving destination, and the surrounding environment information at least includes the driving environment data of roads, vehicles, pedestrians, and traffic signals; the built-in camera and voice equipment serve as the communication medium between the target driver and the vehicle control system;

[0063] S12, using data management software to input the collected driving data into the system background; the driving data includes the basic data, the driving environment data and the behavior data.

[0064] In the specific implementation process:

[0065] When the target driver starts the vehicle, the vehicle control system will first read and confirm the basic information such as the license plate number, vehicle model, and driver's license, and obtain the current location information of the vehicle in real time through the GPS positioning system. The target driver sets the driving destination through the vehicle navigation system or voice command, and the system records the acquired information and uses it as part of the basic data; the external Tbox and radar system work together to collect real-time road information around the vehicle, including road type (such as highways, city streets, etc.), road width, curvature, etc., as well as dynamic information such as the location and speed of surrounding vehicles and pedestrians to provide the necessary basis for obstacle avoidance; by identifying the color and status of traffic lights and the instructions of traffic signs, the intelligent driving assistance system determines the current traffic conditions; the built-in camera captures the target driver's facial expressions and body language in real time, such as fatigue, distraction, tension, etc., to evaluate the target driver's attention level and driving safety, and receives and recognizes the target driver's voice commands and conversation content in real time through the built-in voice recognition device;

[0066] The vehicle control system receives and stores data from the vehicle control system, external Tbox, radar, built-in camera, voice equipment and other devices in real time through the data management software; according to the nature and purpose of the data, the data management software divides the collected driving data into three categories: basic data, driving environment data and behavior data, and stores them in the system separately.

[0067] For S2, specifically:

[0068] S21, based on the existing assisted driving function, simulating a driving process of a preset vehicle and performing a driving test on the driving process;

[0069] S22, collecting the test surrounding environment information in real time during the driving test through the sensors of the preset vehicle;

[0070] S23, identifying a test driving scenario based on the test surrounding environment information;

[0071] S24, according to the driving scenario, the system simulates the behavior pattern of each intelligent agent;

[0072] S25, performing a driving test on the behavior pattern of each intelligent agent in each of the driving scenarios through a simulation experiment or an actual road;

[0073] S26, during the driving test, recording the behavior indicators of each intelligent agent to evaluate the functional effectiveness and safety, and obtaining an evaluation result;

[0074] S27, establishing an initial multi-intelligent system based on the evaluation results, including a driving agent, an automatic planning agent, a perception agent and a core control agent.

[0075] In the specific implementation process:

[0076] Simulate the driving process of a preset vehicle:

[0077] A simulated driving environment is established, and existing assisted driving functions such as automatic cruise control and lane keeping are used to simulate the driving of vehicles on actual roads for driving tests. During the simulated driving process, environmental data is collected in real time through the preset vehicle sensors. Based on the collected environmental information, the current driving scene is identified, such as highways, city streets, intersections, curves, pedestrian dynamics, etc. For each identified driving scene, the system simulates the behavior patterns of each agent according to the preset logic. The agents include driving agents (responsible for the basic driving operations of the vehicle), automatic planning agents (responsible for route planning and decision-making), perception agents (responsible for environmental perception and information processing) and core control agents (responsible for Coordination and control of the overall system); in a simulated environment, a driving test is performed on the behavior pattern of each agent in each driving scenario. During the test, the system records the behavior indicators of each agent, such as response time, decision accuracy, safety, etc. Based on the recorded behavior indicators, it is analyzed whether each agent can correctly identify and respond to changes in the driving scenario during the test, and whether its behavior meets the safety standards, so as to evaluate the functional effectiveness and safety of each agent's behavior pattern to establish an initial multi-agent system; based on the evaluation results, the agent's behavior pattern is adjusted and optimized to establish an initial multi-agent system including a driving agent, an automatic planning agent, a perception agent, and a core control agent.

[0078] Among them, for S22, specifically:

[0079] S221, collecting surrounding environment data of the driving test process in real time through the sensors of the preset vehicle;

[0080] S222, automatically screening the factors of the surrounding environment data through Rule-based RewardModel to obtain the final test surrounding environment information, wherein the factor automatic screening rules at least include screening out information threatening driving safety and information related to traffic rules through the existing auxiliary driving system; the factor automatic screening rules are automatically adjusted according to the changes in the driving environment and the update of sensor data according to the new environmental conditions and task requirements.

[0081] In the specific implementation process:

[0082] During the driving test, the vehicle's preset sensors (such as cameras, Tbox, radar, etc.) collect environmental data around the vehicle in real time. The environmental data includes but is not limited to road conditions, traffic signals, the location and speed of other vehicles, pedestrians, obstacles, etc.

[0083] Automatic factor screening based on Rule-based RewardModel:

[0084] The collected environmental data is filtered through the rule-based reward model. According to the preset rules, the information useful for the test process is filtered out from the raw data. The screening rules include at least information that threatens driving safety and information related to traffic rules. Information that threatens driving safety, such as obstacles that suddenly appear in front, sudden lane changes of vehicles, etc., is filtered out through existing auxiliary driving systems (such as collision warning systems, lane departure warning systems, etc.); information related to traffic rules involves compliance with traffic regulations, such as red light stop, green light go, speed limit signs, traffic police gestures, etc. The system needs to recognize and understand traffic signs and signals to ensure that the test vehicle complies with traffic rules during the test; and the screening rules will automatically adjust the screening rules according to new environmental conditions and task requirements. For example, when driving on the highway, the system needs to pay more attention to the speed and distance of the vehicle; while when driving on city streets, the system needs to pay more attention to the dynamics of pedestrians and non-motor vehicles; after automatic factor screening, the surrounding environment information set required in the test driving scene is obtained.

[0085] For S3, specifically:

[0086] S31, establish the behavior state factors and list them as the Q-table of the agent operation;

[0087] S32, initializing the Q-table, initializing the Q value to 0, and performing driving training based on the improved Q-learning algorithm, the formula of which is:

[0088]

[0089] Among them, γ (reward), s (state), a (action), among which σ 2 is the influence of the factor of the selected agent, the initial value is 1, and the square exploration method can be used to adjust the optimal factor. It is the adjustment of the core control body;

[0090] S33, the preset vehicle enters the test process, and according to the current vehicle state s of the preset vehicle, a driving action a1 of a preset agent is selected and operated, and then the preset agent selects driving action a2 to operate, and obtains feedback results after a1 and a2 operations; wherein the parameter of the agent parameter a1 is σ 2 ;

[0091] S34, comparing the feedback result with the target driver behavior, the best behavior s' and the reward γ, and updating to the following formula Q(S, a);

[0092] Q(S,a)=Q(S,a)+α[R(S,a)+γmaxQ′(s′,a′)-Q(S,a)

[0093] S35, executing the test process one by one for each agent in each scene until the data in the Q-table is updated, obtaining the optimal action path in each driving state, and updating the optimal action path to the multi-agent system.

[0094] In the specific implementation process:

[0095] Reference Figure 3 , is a schematic diagram of a Q-table for calculating expected returns in one embodiment of the present invention, including:

[0096] A behavior state factor is established and listed as a Q-table of the agent's operation; the behavior state factor is used to describe the characteristics or attributes of the agent of the preset vehicle under different driving states, and the Q-table is used to store the expected rewards (Q values) obtained by the agent taking different operations under different states, each row represents a state, and each column represents an operation.

[0097] Initialize the Q-table and initialize all Q values ​​to 0, indicating that the agent has no prior knowledge of any state and action at the beginning. The following formula is an improved variant of the Q-learning algorithm, which contains some optimizations and adjustments to better adapt to the specific needs of driving training. The formula is:

[0098]

[0099] Where s is the state, a is the action, γ is the reward, σ 2 is the influence of the factors of the selected agent, It is the adjustment of the core control body; wherein, the influence of the factors of the selected intelligent agent refers to the weights of certain factors (such as safety, efficiency, etc.) when selecting actions. These weights can be adjusted by methods such as the square exploration method to find the optimal factor combination.

[0100] The preset vehicle enters the test process. The current driving state of the preset vehicle is s. Select driving action a1 and perform the operation. The selection of driving action is that the agent selects an action (such as acceleration, deceleration, steering, etc.) according to the current state and Q-table and enters the test. After selecting driving action a1 for testing, select driving action a2 under the same agent for operation and testing. After executing the action, the system will give feedback, including new status, rewards, etc., and compare the feedback results with the target driver behavior, the best behavior s' and the reward, update the Q value, and compare the actual feedback with the target driver behavior to evaluate the quality of the action; use the Q-learning update formula to update the Q value in the Q-table according to the feedback results and rewards. The update formula is:

[0101] Q(S,a)=Q(S,a)+α[R(S,a)+γmaxQ′(s′,a′)-Q(S,a)

[0102] The test process is executed one by one for each agent in each scene until the data in the Q-table is updated. After multiple iterative training, the Q value in the Q-table will gradually stabilize, reflecting the strategy of taking the best action in different states. According to the final stable Q-table, the best action path in each driving state is obtained, and the best action path is applied to the multi-agent system.

[0103] For S4, specifically:

[0104] S41, determining whether to enter the driving mode;

[0105] S42, if entering the driving mode, triggering the perception agent to obtain the current driving state factor;

[0106] S43, based on the current driving state factor, obtaining an optimal driving operation path in the current state according to the multi-agent system;

[0107] S44, transmitting the optimal driving operation behavior instruction to each intelligent agent through the core intelligent agent;

[0108] Determining whether to execute the optimal driving operation path based on the current behavior state of the target driver;

[0109] S45: If the optimal driving operation path is executed, the driving operation is performed according to the program of the current optimal driving operation path.

[0110] In the specific implementation process:

[0111] The system determines whether the vehicle has entered the driving mode through data input from multiple sensors (such as door switches, seat belt status, ignition status, etc.). Once entering the driving mode, the system will trigger the perception agent to collect the current driving state factors, including but not limited to vehicle speed, lane position, surrounding vehicles and pedestrians, traffic signal status, etc.; based on the current driving state factors, the system uses a multi-agent system to analyze the current driving state factors and match the best driving operation path under the given state, which may include steering, acceleration, deceleration, lane changing and other operations; the best driving operation behavior instructions are transmitted to each agent through the core agent; the current target driver's behavior state is identified through sensors. If the target driver has abnormal behavior (such as whether he is concentrating, whether he has signs of fatigue, etc.), it is determined that the best driving operation path needs to be executed. When the system determines that the best driving operation path should be executed, the driving operation is performed according to the program of the current best driving operation path.

[0112] Reference Figure 2 , the structural block diagram of the driving assistance method system in the present invention includes:

[0113] A data processing module, used to detect and identify the surrounding environment information of a preset vehicle and the behavior information of a target driver, analyze and process the surrounding environment information and the behavior information of the target driver, and extract driving data that affects driving operations;

[0114] The system establishment module is used to simulate the preset vehicle driving process and establish the initial multi-agent system based on the existing assisted driving function;

[0115] A system training module, which performs driving training based on the target driver on the initial multi-agent system according to an improved Q-learning algorithm, obtains an optimal action path and updates it to the multi-agent system;

[0116] The operation execution module directly transmits the optimal action path to each intelligent agent and executes the driving operation of the optimal action path according to the current driving state.

[0117] Reference Figure 4 In an embodiment of the present invention, a computer device is also provided. The computer device may be a server, and its internal structure may be as follows: Figure 4As shown. The computer device includes a processor, a memory, a display screen, an input device, a network interface and a database connected through a system bus. Among them, the processor designed by the computer is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store the corresponding data in this embodiment. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, the above method is implemented.

[0118] Those skilled in the art will understand that Figure 4 The structure shown in the figure is merely a block diagram of a portion of the structure related to the solution of the present invention, and does not constitute a limitation on the computer device to which the solution of the present invention is applied.

[0119] An embodiment of the present invention further provides a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, the above method is implemented. It can be understood that the computer-readable storage medium in this embodiment can be a volatile readable storage medium or a non-volatile readable storage medium.

[0120] In summary, a driving assistance method detects and identifies the surrounding environment information of a preset vehicle and the behavior information of a target driver, analyzes and processes the surrounding environment information and the behavior information of the target driver, and extracts driving data that affects driving operations; based on the existing assisted driving function, simulates the driving process of the preset vehicle and establishes an initial multi-agent system; based on an improved Q-learning algorithm, the initial multi-agent system is trained to drive based on the target driver, and an optimal action path is obtained and updated to the multi-agent system; the optimal action path is directly transmitted to each agent, and the driving operation of the optimal action path is executed according to the current driving state, so as to realize customized intelligent driving assistance methods for target drivers, by identifying specific target drivers, understanding their driving habits for a long time, strengthening the ability of intelligent driving, standardizing and optimizing the driving behavior of target drivers, reducing driving energy consumption, and improving driving safety.

[0121] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media provided by the present invention and used in the embodiments can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double-speed data rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM.

[0122] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, device, article or method including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, device, article or method. In the absence of further restrictions, an element defined by the sentence "includes a ..." does not exclude the presence of other identical elements in the process, device, article or method including the element.

[0123] The above description is only a preferred embodiment of the present invention, and does not limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made by using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A driving assistance method, characterized in that: Detect and identify the surrounding environment information of the preset vehicle and the behavior information of the target driver, analyze and process the surrounding environment information and the behavior information of the target driver, and extract driving data that affects the driving operation; Based on the existing assisted driving functions, simulate the preset vehicle driving process and establish the initial multi-agent system; Based on the improved Q-learning algorithm, the initial multi-agent system is trained based on the target driver to obtain the optimal action path and update it to the multi-agent system; The optimal action path is directly transmitted to each intelligent agent, and the driving operation of the optimal action path is executed according to the current driving state.

2. The driving assistance method according to claim 1, characterized in that: The steps of detecting and identifying the surrounding environment information of a preset vehicle and the target driver's behavior information, analyzing and processing the surrounding environment information and the target driver's behavior information, and extracting driving data that affects the driving operation include: The basic data of the preset vehicle is loaded through the vehicle control system, the peripheral environment information of the preset vehicle is collected automatically through the external Tbox and radar of the preset vehicle, and the behavior data of the target driver is collected through the built-in camera and voice equipment of the preset vehicle; the basic data at least includes the license plate number, vehicle model, driver's license, the location of the preset vehicle, and the driving destination, and the surrounding environment information at least includes the driving environment data of roads, vehicles, pedestrians, and traffic signals; the built-in camera and voice equipment serve as the communication medium between the target driver and the vehicle control system; The collected driving data is entered into the system background using data management software; the driving data includes the basic data, the driving environment data and the behavior data.

3. The driving assistance method according to claim 1, characterized in that: The steps of simulating a preset vehicle driving process based on the existing assisted driving function and establishing an initial multi-agent system include: Based on the existing assisted driving function, simulate the driving process of a preset vehicle and perform a driving test on the driving process; Collecting the test surrounding environment information in real time during the driving test through the sensors of the preset vehicle; Based on the test surrounding environment information, identifying a test driving scenario; According to the driving scenario, the system simulates the behavior patterns of each agent; Conducting driving tests on the behavior patterns of each intelligent agent in each of the driving scenarios through simulation experiments or actual roads; During the driving test, the behavior indicators of each intelligent agent are recorded to evaluate the functional effectiveness and safety, and an evaluation result is obtained; An initial multi-intelligent system is established based on the evaluation results, including a driving agent, an automatic planning agent, a perception agent and a core control agent.

4. The driving assistance method according to claim 3, characterized in that: The step of collecting the test surrounding environment information in real time during the driving test by using the sensor of the preset vehicle includes: Collecting the surrounding environment data of the driving test process in real time through the sensors of the preset vehicle; The surrounding environment data is automatically screened by factors through the Rule-based Reward Model to obtain the final test surrounding environment information, wherein the factor automatic screening rules at least include screening out information that threatens driving safety and information related to traffic rules through the existing auxiliary driving system; the factor automatic screening rules are automatically adjusted according to the changes in the driving environment and the update of sensor data according to new environmental conditions and task requirements.

5. The driving assistance method according to claim 1, characterized in that: The step of performing driving training based on the target driver on the initial multi-agent system based on the improved Q-learning algorithm to obtain the optimal action path and update it to the multi-agent system includes: Establish behavioral state factors and list them as Q-tables for agent operations; Initialize the Q-table, initialize the Q value to 0, and perform driving training based on the improved Q-learning algorithm, the formula of which is: Among them, γ (reward), s (state), a (action), among which σ 2 is the influence of the factor of the selected agent, the initial value is 1, and the square exploration method can be used to adjust the optimal factor. It is the adjustment of the core control body; The preset vehicle enters the test process, and according to the current vehicle state s of the preset vehicle, a driving action a1 of a preset agent is selected and operated, and then the preset agent selects driving action a2 to operate, and obtains feedback results after a1 and a2 operations; wherein the parameter of the agent parameter a1 is σ 2 ; Compare the feedback result with the target driver behavior, the best behavior s' and the reward γ, and update it to the following formula Q(S, a); Q(S,a)=Q(S,a)+α[R(S,a)+γmaxQ′(s′,a′)-Q(S,a) The test process is executed one by one for each agent in each scene until the data in the Q-table is updated, the optimal action path in each driving state is obtained, and the optimal action path is updated to the multi-agent system.

6. The driving assistance method according to claim 1, characterized in that: The step of directly transmitting the optimal action path to each agent and executing the driving operation of the optimal action path according to the current driving state includes: Determine whether to enter the driving mode; If the driving mode is entered, the perception agent is triggered to obtain the current driving state factor; Based on the current driving state factor, obtaining an optimal driving operation path in the current state according to the multi-agent system; The optimal driving operation behavior instruction is transmitted to each intelligent agent through the core intelligent agent; Determining whether to execute the optimal driving operation path based on the current behavior state of the target driver; If the optimal driving operation path is executed, the driving operation is performed according to the procedure of the current optimal driving operation path.

7. A driving assistance method system, characterized in that: include: A data processing module, used to detect and identify the surrounding environment information of a preset vehicle and the behavior information of a target driver, analyze and process the surrounding environment information and the behavior information of the target driver, and extract driving data that affects driving operations; The system establishment module is used to simulate the preset vehicle driving process and establish the initial multi-agent system based on the existing assisted driving function; A system training module, which performs driving training based on the target driver on the initial multi-agent system according to an improved Q-learning algorithm, obtains an optimal action path and updates it to the multi-agent system; The operation execution module directly transmits the optimal action path to each intelligent agent and executes the driving operation of the optimal action path according to the current driving state.

8. A computer device comprising a memory and a processor, wherein a computer program is stored in the memory, wherein: When the processor executes the computer program, the steps of the driving assistance method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the driving assistance method according to any one of claims 1 to 6 are implemented.

Citation Information

Patent Citations

  • Driving distraction identification method and system under multiple scenes and vehicle

    CN114943956A

  • Vehicle anthropomorphic decision control method and device, vehicle and storage medium

    CN115923833A

  • Driving aid decision-making method and system based on vehicle-mounted chip

    CN118312750A

  • Automobile automatic cruise driving assistance system and using method thereof

    CN118323191A

  • Auxiliary driving method and system and storage medium

    CN118387132A

Cited By

  • Engineering truck safe driving monitoring system based on face and behavior recognition fusion

    CN119888696A

  • A safe driving monitoring system for engineering vehicles based on the fusion of facial and behavioral recognition

    CN119888696B