Self-adaptive cruise safety self-evolution method, device and system based on safety boundary

Abstract

The invention belongs to the technical field of vehicle driving control, and particularly relates to an adaptive cruise safety self-evolution method, device and system based on a safety boundary. The method comprises the steps that vehicle driving data and environment information are collected and preprocessed, and a time sequence data set containing the state, control input and corresponding reward values of a vehicle in all time steps is obtained; secondly, dynamically evaluating the safety distance between the vehicle and surrounding obstacles through a control obstacle function, determining a safety area, and constructing and updating a safety driving boundary function of the vehicle at the same time; the MPC controller takes a safe driving boundary function as a constraint condition and takes a stable approaching target speed and a vehicle following distance as targets to obtain a control input sequence; and the vehicle executes corresponding operation according to the control parameters, and continuously adjusts the weight matrix of the MPC through an online optimization algorithm, so that the control parameters are gradually optimized. According to the method, adaptive adjustment of the MPC parameters is realized, so that the safety and the adaptive capacity of the vehicle in a complex traffic environment are improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of vehicle driving control, and in particular relates to a safety boundary-based adaptive cruise safety self-evolution method, device and system. Background Art

[0002] As an important part of intelligent driving, adaptive cruise control provides a more comfortable and safe driving experience by adjusting vehicle speed and following distance. However, the existing adaptive cruise system still has limitations in complex traffic environments, especially in the face of sudden traffic conditions. The safety and reliability of the system still needs to be further improved.

[0003] In order to solve the challenges faced by adaptive cruise control in real environments, it is necessary to introduce a self-optimization module with a data closed loop. By continuously collecting and analyzing real-time data, the system can gradually adjust its safety boundaries, control strategies, and perception algorithms. The core of this process lies in the continuous accumulation and learning of data, allowing the system to adapt and optimize under different driving conditions. The data closed loop can not only improve the safety of the intelligent driving system, but also realize its self-evolution ability in complex traffic scenarios. In addition, with the continuous development of intelligent driving technology, safety has become an important research direction. Through the data closed loop, the system can identify potential risks in a timely manner and make corresponding adjustments, thereby maintaining a high level of safety performance under various environmental conditions.

[0004] In this context, some new technical solutions have emerged. Patent CN116300850A proposes a human-like safety self-evolution method and system for autonomous driving based on data mechanism fusion. This method constructs a driver's decision-making and planning model by extracting real human driving characteristics. However, this method may face some limitations in practical applications, including a high dependence on historical data and insufficient adaptability to complex traffic scenarios, which makes it less flexible in dynamic and changeable driving environments. Patent CN117609099A provides a general self-evolution method and system for autonomous driving algorithms. Although it improves algorithm performance through adversarial scenario generation and interactive training, it may still encounter overfitting problems in complex environments, that is, the algorithm performs well in specific scenarios, but cannot effectively respond to other unknown scenarios. Therefore, although it improves the intelligence and performance boundaries of the algorithm, it still cannot guarantee reliability in various dynamic and open traffic environments. Summary of the invention

[0005] The purpose of the present invention is to provide a safety boundary-based adaptive cruise safety self-evolution method, device and system, which combines the control obstacle function with reinforcement learning, realizes adaptive adjustment of the controller parameters of the model predictive control (MPC) through real-time data feedback and dynamic optimization control strategy, thereby improving the safety and adaptability of the system in complex traffic environments.

[0006] To achieve the above object, the present invention provides an adaptive cruise safety self-evolution method based on safety margin, comprising the following steps:

[0007] S1. Real-time collection and preprocessing of vehicle driving data and environmental information to obtain the state x of the vehicle at each time step. t , control input sequence u t And the corresponding reward value R t Time series dataset D;

[0008] S2. According to the time series data set D, dynamically evaluate the safe distance between the vehicle and surrounding obstacles by controlling the obstacle function, determine the safe area, and construct and update the vehicle's safe driving boundary function h(x);

[0009] S3, MPC controller uses the safe driving boundary function h(x) as the constraint condition and aims to smoothly approach the target speed and following distance to obtain the control input sequence u t ;

[0010] S4: The vehicle performs corresponding operations according to the control parameters. During the execution, an online optimization algorithm is used to maximize the reward value R. t The long-term cumulative value of is taken as the target, and the weight matrix of the MPC controller is continuously adjusted to gradually optimize the control parameters.

[0011] Furthermore, in step S1, the vehicle driving data includes the current vehicle speed v t , acceleration a t , distance to the vehicle ahead d t The environmental information includes road type, road status, traffic flow density, weather conditions, and the distribution and movement information of surrounding dynamic obstacles. Through the comprehensive collection and analysis of these data, accurate input is provided for the vehicle's safety boundary modeling, thereby achieving dynamic optimization of the adaptive cruise strategy and improving safety.

[0012] Furthermore, in step S2, the method for determining the safety area includes: first, constructing an initial control obstacle function through an initial safety controller, and then further optimizing and constructing the safety controller based on the initial control obstacle function by judging whether the vehicle exceeds the lane boundary line, and then regenerating a new control obstacle function; then, the system checks whether the current safety area can continue to expand, and if so, repeats the process to increase the safety area; if not, stops expanding, and outputs the final control obstacle function result to ensure that the system maximizes the safety area while meeting the safety boundary.

[0013] Furthermore, the constraint condition refers to a safe driving boundary function h(x)≥0; wherein, when the vehicle is within the safety area, h(x)≥0, and when the vehicle is not within the safety area, h(x)<0.

[0014] Furthermore, in step S3, the objective function of the MPC controller is:

[0015]

[0016] Among them, L(x(t),u(t))=‖x(t)-x target ‖ 2 +R‖u(t)‖ 2 is the stage cost function, which indicates the deviation between the current state of the vehicle and the target state; x target is the expected cruising state, preferably including the target speed and safety distance; R is the weight matrix of the control input; u(t) is the control input sequence;

[0017] Φ(x(t+N))=Q‖x(t+N)-x goal ‖ 2 is the terminal cost function, which is used to evaluate the final state of the system at the end of the prediction time domain; where x goal represents the ideal final state, preferably including maintaining a safe distance from the vehicle in front and the desired speed; Q is the weight matrix of the state vector, which is used to balance the importance of the terminal state.

[0018] Furthermore, in step S4, the online optimization algorithm uses deep Q learning to calculate the reward value R at each time step. t , obtain the long-term cumulative value of the reward value, and train with the goal of maximizing the long-term cumulative value of the reward value so that the control parameters gradually converge to the optimal strategy;

[0019] Among them, the reward value R t The formula is as follows:

[0020]

[0021] Furthermore, the state space S of the deep Q learning is S={x t ,u t ,E}, action space A={u min ,u max}, where E represents the surrounding environment information; u min Indicates the minimum speed, u max represents the maximum speed; the iterative update rule adopted by the deep Q learning is as follows:

[0022]

[0023] Among them, s t Indicates the current state, s t+1 Indicates the state at the next moment, a t represents the action taken at the current moment, R refers to the immediate reward, α is the learning rate, and γ is the discount factor.

[0024] The present invention also provides an adaptive cruise safety self-evolution device based on a safety boundary, which adopts any of the above adaptive cruise safety self-evolution methods, including:

[0025] A real-time data feedback and learning module, for organizing vehicle driving data and environmental information into a time series data set D, and continuously adjusting the weight matrix of the MPC controller through an online optimization algorithm so that the control parameters are gradually optimized;

[0026] A safety boundary calculation module is used to dynamically evaluate the safety distance between the vehicle and surrounding obstacles by controlling the obstacle function according to the time series data set D, and to construct and update the vehicle's safe driving boundary function h(x);

[0027] An MPC controller module is used to transmit control parameters to the safety margin calculation module for correction to obtain safe control parameters according to the current vehicle state and the prediction of the future state;

[0028] The execution module is used to execute the safe control parameters obtained by the MPC controller module to adjust the actual driving state of the vehicle.

[0029] The present invention also provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the adaptive cruise safety self-evolution method described in any one of the above items is implemented.

[0030] The present invention also provides an adaptive cruise safety self-evolution system based on safety boundaries, comprising a computer-readable storage medium and a processor;

[0031] The computer-readable storage medium is used to store executable instructions;

[0032] The processor is used to read the executable instructions stored in the computer-readable storage medium and execute any of the above methods.

[0033] In general, the above technical solution conceived by the present invention has the following technical advantages compared with the prior art:

[0034] 1. The safety boundary-based adaptive cruise safety self-evolution method provided by the present invention aims at the problem of insufficient safety and adaptability of existing adaptive cruise control systems in dynamic traffic environments. It combines the control obstacle function with reinforcement learning, and realizes adaptive adjustment of the model predictive control (MPC) controller parameters through real-time data feedback and dynamic optimization control strategy, thereby improving the safety and adaptability of vehicles in complex traffic environments. Compared with the prior art, the present invention uses real-time data feedback to continuously optimize and adjust the algorithm, so that it has higher adaptability in complex and changeable traffic scenes.

[0035] 2. The present invention emphasizes the safety of the method. It calculates the safety boundary through CBF (control barrier function), evaluates the safety status of the vehicle in real time, and ensures that the vehicle is always in a safe driving range through the defined safety boundary function h(x) to avoid potential collision risks; it ensures the safety of the control parameters by receiving and correcting the control parameters of MPC. Combined with reinforcement learning to optimize the MPC control parameters, through continuous iteration and optimization, it can quickly respond to emergencies and reduce accident risks, thereby realizing intelligent control of the vehicle's driving status and ensuring safety under various driving conditions.

[0036] 3. The MPC controller of the present invention is based on the dynamic model of the vehicle, combined with the safety boundary constraints and the target state, and generates the cruise command by solving the optimal control input u(t). The MPC controller generates the control input sequence u(t) using the stage cost function and the terminal cost function, and ensures that the vehicle approaches the target state smoothly under the condition of satisfying the safety boundary constraints. At the same time, the newly collected data is updated to the historical data set through the real-time data feedback and learning module to form a complete optimization chain, which promotes the MPC parameters to adapt to dynamic environmental changes. At the same time, the reinforcement learning (RL) algorithm is integrated to optimize the parameter settings of the MPC controller by analyzing the real-time feedback data, so that the system can adapt to different traffic environments and driving conditions, and continuously improve the safety and adaptability of the adaptive cruise system.

[0037] 4. The key to the present invention lies in the data closed loop of real-time data feedback and learning modules. By continuously collecting and analyzing real-time data, the vehicle can quickly respond safely under different driving conditions and reduce the risk of accidents. Through this module, the present invention can effectively improve the reliability and safety performance of the intelligent driving system, and provide a solid theoretical foundation and practical guarantee for the future development of autonomous driving technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 A safety self-evolution flowchart of adaptive cruise safety based on safety boundary provided by the present invention;

[0039] Figure 2 Calculation flow chart of the safety boundary. DETAILED DESCRIPTION

[0040] In order to make the purpose, technical scheme and advantages of the present invention clearer, the present invention is further described in detail below in conjunction with the 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. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0041] See also Figure 1 and 2 The present invention provides an adaptive cruise safety self-evolution method based on safety margin, aiming to solve the problems of insufficient safety and adaptability faced by existing adaptive cruise systems in dynamic traffic environments. The method introduces a control barrier function (CBF) to calculate the safety margin and combines reinforcement learning (RL) to optimize the model predictive control (MPC) parameters, thereby ensuring that the vehicle can achieve safe and efficient adaptive cruise under complex and changeable driving conditions.

[0042] In order to realize the adaptive cruise safety self-evolution, the present invention first provides an adaptive cruise safety self-evolution system, which is mainly composed of four modules, including a real-time data feedback and learning module, a safety boundary calculation module, an MPC controller and an execution module. When the system enters the adaptive cruise mode, the real-time data feedback and learning module first collects the vehicle's driving data (including speed, acceleration, distance to the front vehicle, traffic signals, etc.) and environmental information in real time through sensors on the vehicle (such as radar, camera, lidar, GPS, etc.). These data are organized into a time series data set D, which contains the state x of the vehicle at each time step. t , control input u t And the corresponding reward value R t, as the input of the reinforcement learning (RL) algorithm. At the same time, the preprocessed data is input to the safety boundary calculation module, and the trained MPC control parameters and preprocessed data are transmitted to the MPC controller.

[0043] The safety margin calculation module uses the control obstacle function (CBF) theory to dynamically evaluate the safe distance between the vehicle and surrounding obstacles based on the received pre-processed data, and constructs and updates the vehicle's safe driving boundary function h(x). This safety boundary function ensures that the vehicle always stays within the safe range during driving to avoid potential collision risks. The safety margin calculation module is also responsible for receiving the control parameters of the MPC and making corrections to ensure the safety of the control parameters.

[0044] Subsequently, the MPC controller generates cruise instructions by solving the optimal control input u(t) based on the current vehicle state and the prediction of the future state, combined with the safety boundary constraints (i.e., the control parameters are transmitted to the safety boundary calculation module for correction) and the target state. The MPC controller generates the control input sequence u(t) using the stage cost function and the terminal cost function, and ensures that the vehicle approaches the target speed and following distance smoothly under the safety boundary constraints. These control instructions are then sent to the execution module and executed by the vehicle's actuators (such as the accelerator and braking system) to adjust the actual driving state of the vehicle.

[0045] During the execution of control instructions, the real-time data feedback and learning module continuously receives the latest driving status of the vehicle. The real-time data feedback and learning module uses these feedbacks to adjust the weight matrices R and Q in the MPC controller through an online optimization algorithm, so that the control strategy can better adapt to different traffic environments and driving conditions. With the continuous iteration and update of the algorithm, the real-time data feedback and learning module promotes the MPC parameters to gradually converge to the optimal configuration, enhances the system's adaptability in complex dynamic environments, and ensures the safety of evolution.

[0046] Among them, MPC calculates the optimal control input sequence within a limited time range in the future by optimizing the objective function. At the current moment, based on the system state and control input, MPC uses the model to predict the system response within a period of time in the future. This future period is called the prediction domain or prediction interval. The weight matrices Q and R are key parameters in the MPC cost function, representing the weights of state variables and control inputs, respectively, and are used to adjust control performance and robustness. In practical applications, the determination of Q and R matrices usually depends on the system model and design requirements, and optimization algorithms can be used to adjust the weights of the matrices to achieve the desired control performance.

[0047] The motion control algorithm of the present invention provides a self-evolving module through dynamic adjustment of the safety margin and real-time optimization of MPC, so that the algorithm can adapt to different driving environments and traffic conditions. During the adaptive cruise process, the algorithm can not only learn from historical data, but also provide real-time feedback based on current traffic conditions to continuously improve the safety and response capabilities of the system.

[0048] The adaptive cruise safety self-evolution method provided by the present invention comprises the following steps:

[0049] S1. Real-time collection and preprocessing of vehicle driving data and environmental information to obtain the state x of the vehicle at each time step. t , control input sequence u t And the corresponding reward value R t Time series dataset D;

[0050] S2. According to the time series data set D, dynamically evaluate the safe distance between the vehicle and surrounding obstacles by controlling the obstacle function, determine the safe area, and construct and update the vehicle's safe driving boundary function h(x);

[0051] S3, MPC controller uses the safe driving boundary function h(x) as the constraint condition and aims to smoothly approach the target speed and following distance to obtain the control input sequence u t ;

[0052] S4: The vehicle performs corresponding operations according to the control parameters. During the execution, an online optimization algorithm is used to maximize the reward value R. t The long-term cumulative value of is taken as the target, and the weight matrix of the MPC controller is continuously adjusted to gradually optimize the control parameters.

[0053] In step S1, the vehicle driving data includes the current vehicle speed v t , acceleration a t , distance to the vehicle ahead d t and traffic signals. The environmental information includes road type, road status, traffic flow density, weather conditions, and the distribution and movement information of surrounding dynamic obstacles. Through the comprehensive collection and analysis of these data, accurate input is provided for the vehicle's safety boundary modeling, thereby achieving dynamic optimization of the adaptive cruise strategy and safety improvement.

[0054] Specifically, the present invention uses reinforcement learning to optimize the parameters of the MPC controller to improve the safety and adaptability of the system. The real-time data feedback and learning module continuously collects the vehicle's driving data, including the current speed v t , acceleration a t and the distance d to the vehicle in front tThe weight matrices R and Q in the MPC controller are dynamically adjusted through deep Q learning (DQN), thereby optimizing the control strategy and improving the system's adaptability and response speed. The real-time data feedback and learning module learns how to dynamically adjust the weight matrices R and Q in the MPC controller to adapt to different traffic scenarios and driving conditions by constantly interacting with the environment. During the training process, DQN receives a reward signal R t , optimize the strategy to maximize the long-term cumulative reward, thereby improving the overall performance of the system. The state space S and action space A of DQN are defined as: S = {x t ,u t ,E},A={u min ,u max}, the state space S contains the current vehicle state x t , control input u t and surrounding environment information E, while the action space A includes the possible acceleration and deceleration range of the vehicle, u min Indicates the minimum speed, u max In order to guide the learning process, the reward function R is set to reflect the safety and efficiency of the system. The specific form is as follows:

[0055]

[0056] Among them, h(x) is the safety boundary function. h(x) ≥ 0 means that the vehicle is in a safe driving range and obtains positive rewards to enhance the stability of the system. h(x) < 0 means that the vehicle is approaching an obstacle or entering an unsafe area. The system gives negative rewards, prompting the reinforcement learning algorithm to quickly adjust the control strategy to avoid potential dangers. This reward module enables the reinforcement learning algorithm to focus more on optimizing the safety range and avoiding collisions or other dangerous situations. To improve learning efficiency, the system uses the Q value update formula:

[0057]

[0058] Among them, s t Indicates the current state, s t+1 Indicates the state at the next moment, a t represents the action taken at the current moment, R refers to the immediate reward, α is the learning rate, and γ is the discount factor.

[0059] Through continuous iterative updates, the Q value can converge to the optimal strategy, thereby improving the vehicle's adaptive ability. At each time step, the system monitors the vehicle's status and environmental information through sensors to form a new data set D. This data set records the status, action and corresponding rewards at each moment, which is helpful for subsequent learning and optimization. The newly collected data will be continuously updated to the historical data set for use in subsequent learning, so that the reinforcement learning algorithm can effectively utilize the latest data, thereby improving the accuracy and efficiency of learning. The real-time data feedback and learning module ensure that the parameters of the MPC controller can be dynamically adjusted according to the real-time traffic environment and vehicle status, so as to achieve efficient, safe and comfortable operation of the adaptive cruise system in complex traffic scenarios. With the continuous optimization of the reinforcement learning algorithm and the dynamic adjustment of parameters, the system can gradually converge to the optimal strategy, significantly improving the intelligence level and self-evolution capability of adaptive cruise control.

[0060] Specifically, the safety margin calculation module of the present invention is based on the mathematical model established by CBF, and can be adaptively adjusted according to the real-time changes of the vehicle and the surrounding environment, so as to ensure that the vehicle does not cross the boundary or collide during cruising, and achieve a dynamic balance between safety and efficiency. The calculation process of the safety margin is as follows: Figure 2 As described. The safety boundary calculation module first constructs an initial CBF through an initial safety controller, and then further optimizes the safety controller based on the initial CBF by judging whether the vehicle exceeds the safety boundary line, and then regenerates a new CBF. Then, the system checks whether the current safety area can continue to expand. If so, the process is repeated to increase the safety area; if not, the expansion is stopped, and the final CBF result is finally output to ensure that the system maximizes the safety area while satisfying the safety boundary. The system state is x(t), and the safety boundary is defined as a differentiable function h(x). When h(x)≥0, the system is in a safe state, and when h(x)<0, the system is in an unsafe state. The safety state of the vehicle is calculated using CBF theory. The safety boundary not only takes into account the motion state of the vehicle, but also integrates the surrounding traffic environment factors, such as the position and speed of other vehicles, pedestrians and obstacles. The calculation formula for the safety boundary is:

[0061]

[0062] in is the derivative of the safety margin, and γ(h(x)) is a non-negative function used to adjust the conservatism of the safety margin. When the traffic environment changes, γ(h(x)) will increase the conservatism of the safety margin, allowing the vehicle to make timely avoidance and deceleration decisions. In a stable driving state, the system reduces conservatism to ensure that the vehicle maintains a high cruising speed, thereby improving efficiency. After the safety margin calculation is completed, the safety margin supervision model predictive control (MPC) module is used as the constraint condition of the MPC controller, and the control parameters of the MPC module are corrected to ensure the driving safety of the vehicle.

[0063] Specifically, after receiving the control parameters corrected by the safety margin calculation module, the MPC controller module of the present invention calculates the optimal control instructions (such as acceleration, braking, etc.) based on the dynamic model and current state of the vehicle to achieve the adaptive cruise control goal. The core goal of the MPC controller is to ensure that the vehicle can not only maintain the set target speed and safety distance, but also achieve smooth acceleration and deceleration control in different driving scenarios by solving the following optimization problems:

[0064]

[0065] Among them, the stage cost function L(x(t),u(t)) = ‖x(t)-x target ‖ 2 +R‖u(t)‖ 2 , represents the deviation between the current state of the vehicle and the target state, and balances the control output to ensure smooth driving; where x target is the expected cruising state (such as the set target speed and safety distance), R is the weight matrix of the control input, which is used to balance the use of the control input, and u(t) is the control input sequence. In the adaptive cruise system, the goal of this cost function is to ensure that the vehicle is close to the target state without generating excessive control output to maintain driving stability.

[0066] Terminal cost function Φ(x(t+N))=Q‖x(t+N)-x goal ‖ 2 It is used to evaluate the final state of the system at the end of the prediction time domain, where x goalrepresents the ideal final state, such as maintaining a safe distance and desired speed from the vehicle in front, and Q is the weight matrix of the state vector, which is used to balance the importance of the terminal state. The MPC controller continuously adjusts the value of u(t) based on the safety boundary information, so that the vehicle always maintains the appropriate target speed and safety distance in a dynamic environment. When the surrounding environment changes (such as the vehicle in front slows down or other obstacles appear), the safety boundary calculation module will update the safety boundary in real time, and the MPC controller will adjust the control input accordingly to ensure that the vehicle achieves the optimal cruising state while maintaining safety. This real-time optimization design enables the adaptive cruise system to always maintain efficient and safe driving performance under different road conditions, vehicle speeds, and traffic densities.

[0067] After receiving the optimal control instructions generated by the MPC controller, the execution module of the present invention is responsible for converting these instructions into actual execution actions of the vehicle, such as acceleration, braking and steering operations. The execution module accurately executes the instructions of the MPC controller through the vehicle's actuators (such as the engine control unit, braking system and steering system) to achieve the target state of adaptive cruise control. This module not only needs to ensure the accurate transmission of control instructions, but also needs to monitor the feedback during the execution process in real time to ensure the effectiveness and response speed of the instructions.

[0068] In summary, the present invention aims at the problem of insufficient safety and adaptability of existing adaptive cruise control systems in dynamic traffic environments, and proposes a safety self-evolution method for adaptive cruise control based on safety margins. The system integrates real-time data feedback and learning modules, safety margin calculation modules, MPC controllers, and execution modules to achieve real-time monitoring, dynamic optimization, and intelligent control of vehicle driving status, thereby significantly improving the performance and reliability of the adaptive cruise control system in complex traffic environments.

[0069] It will be easily understood by those skilled in the art that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A safety self-evolution method for adaptive cruise control based on safety margin, characterized in that: include: S1. Real-time collection and preprocessing of vehicle driving data and environmental information to obtain the state x of the vehicle at each time step. t , control input sequence u t And the corresponding reward value R t Time series dataset D; S2. According to the time series data set D, dynamically evaluate the safe distance between the vehicle and surrounding obstacles by controlling the obstacle function, determine the safe area, and construct and update the vehicle's safe driving boundary function h(x); S3, MPC controller uses the safe driving boundary function h(x) as the constraint condition and approaches the target speed and following distance as the goal to obtain the control input sequence u t ; S4. The vehicle inputs the control sequence u according to the control t Execute the corresponding operation, and during the execution process, use the online optimization algorithm to maximize the reward value R t The cumulative value is the target, and the weight matrix of the MPC controller is continuously adjusted to make the control input sequence u t Optimize step by step.

2. The method for adaptive cruise safety self-evolution based on safety margin according to claim 1 is characterized in that: In step S1, the vehicle driving data includes the current vehicle speed v t , acceleration a t , distance to the vehicle ahead d t and traffic signals, wherein the environmental information includes road type, road status, traffic flow density, weather conditions, and distribution and movement information of surrounding dynamic obstacles.

3. The method for adaptive cruise safety self-evolution based on safety margin according to claim 1 is characterized in that: In step S2, the method for determining the safety area includes: first, constructing an initial control obstacle function through an initial safety controller, then further optimizing and constructing the safety controller based on the initial control obstacle function by judging whether the vehicle exceeds the lane boundary line, and then regenerating a new control obstacle function; then, the system checks whether the current safety area can continue to expand, and if so, repeats the process to increase the safety area; if not, stops expanding, and outputs the final control obstacle function result to ensure that the system maximizes the safety area while meeting the safety boundary.

4. The method for adaptive cruise safety self-evolution based on safety margin according to claim 3 is characterized in that: The constraint condition refers to the safe driving boundary function h(x)≥0; wherein, when the vehicle is within the safety area, h(x)≥0, and when the vehicle is not within the safety area, h(x)<0.

5. The method for adaptive cruise safety self-evolution based on safety margin according to claim 1 is characterized in that: In step S3, the objective function of the MPC controller is: Among them, L(x(t),u(t))=‖x(t)-x target ‖ 2 +R‖u(t)‖ 2 is the stage cost function, which indicates the deviation between the current state of the vehicle and the target state; x target is the expected cruising state, preferably including the target speed and safety distance; R is the weight matrix of the control input; u(t) is the control input sequence; Φ(x(t+N))=Q‖x(t+N)-x goal ‖ 2 is the terminal cost function, which is used to evaluate the final state of the system at the end of the prediction time domain; where x goal represents the ideal final state, preferably including maintaining a safe distance from the vehicle in front and the desired speed; Q is the weight matrix of the state vector, which is used to balance the importance of the terminal state.

6. The method for adaptive cruise safety self-evolution based on safety margin according to claim 5 is characterized in that: In step S4, the online optimization algorithm uses deep Q learning to calculate the reward value R at each time step. t , get the long-term cumulative value of the reward value, and train with the goal of maximizing the long-term cumulative value of the reward value so that the control input sequence u t Gradually converge to the optimal strategy; Among them, the reward value R t The formula is as follows:

7. The method for adaptive cruise safety self-evolution based on safety margin according to claim 6 is characterized in that: The state space S of deep Q learning is S={x t ,u t ,E}, action space A={u min ,u max }, where E represents the surrounding environment information; u min Indicates the minimum speed, u max represents the maximum speed; the iterative update rule adopted by the deep Q learning is as follows: Among them, s t Indicates the current state, s t+1 Indicates the state at the next moment, a t represents the action taken at the current moment, R refers to the immediate reward, α is the learning rate, and γ is the discount factor.

8. An adaptive cruise safety self-evolution device based on safety margin, characterized in that: The method for adaptive cruise safety self-evolution according to any one of claims 1 to 7 comprises: A real-time data feedback and learning module, which is used to organize vehicle driving data and environmental information into a time series data set D, and continuously adjust the weight matrix of the MPC controller through an online optimization algorithm so that the control input sequence is gradually optimized; A safety boundary calculation module is used to dynamically evaluate the safety distance between the vehicle and surrounding obstacles by controlling the obstacle function according to the time series data set D, and to construct and update the vehicle's safe driving boundary function h(x); An MPC controller module is used to transmit the control input sequence to the safety margin calculation module for correction to obtain a safe control input sequence according to the current vehicle state and the prediction of the future state; The execution module is used to execute the safe control input sequence obtained by the MPC controller module to adjust the actual driving state of the vehicle.

9. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and when the computer program is executed by a processor, the adaptive cruise safety self-evolution method as described in any one of claims 1 to 7 is implemented.

10. An adaptive cruise safety self-evolution system based on safety margin, characterized in that: comprising a computer readable storage medium and a processor; The computer-readable storage medium is used to store executable instructions; The processor is used to read the executable instructions stored in the computer-readable storage medium to execute the adaptive cruise safety self-evolution method as described in any one of claims 1-7.

Images