Vehicle body stability control method, system, device, medium and product

Abstract

The present invention belongs to the field of vehicle control and provides a vehicle body stability control method, system, equipment, medium and product. The data such as yaw rate, sideslip angle of center of mass, wheel speed, steering wheel angle, lateral and longitudinal acceleration, driving torque and braking force are obtained by sensors and pre-processed. The control target is then decomposed into multiple sub-problems, steady-state control and weight adjustment are performed, the sub-problems are solved based on the updated weights, and optimization instructions are output, and finally closed-loop feedback is performed for adjustment. In this method, the steady-state weight adaptive algorithm can realize dynamic weight adjustment according to the real-time status of the vehicle and the external environment, and by decomposing multiple objectives into sub-problems and adaptively assigning weights, the on-board multi-core processor is used for parallel solution to shorten the calculation delay, and the weights are gradually adjusted through a flexible weight deletion mechanism, avoiding the impact of drastic fluctuations caused by the weight adjustment process and maintaining the diversity of the population.

Description

Technical Field

[0001] The present invention belongs to the field of vehicle control technology, and specifically relates to a vehicle body stability control method, system, equipment, medium and product. Background Art

[0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.

[0003] As a core technology for vehicle safety, vehicle stability control algorithms, while effective in improving handling and stability, still face shortcomings in multi-objective processing and real-time control decision-making. Vehicle stability control requires simultaneous addressing of multiple objectives, including side slip prevention, anti-lock braking, and traction control. However, these objectives can conflict (for example, during emergency avoidance, stabilizing the vehicle body may conflict with maintaining steering flexibility). Existing vehicle stability control methods typically rely on fixed priority rules, making it difficult to dynamically adjust to actual conditions. Current algorithms capable of performing multi-objective optimization calculations often require significant computational effort, potentially exceeding the computing power of the onboard ECU. This can lead to response delays and compromised control effectiveness.

[0004] Currently, body stability control algorithms have become the preferred configuration for automotive active safety programs, but they still lack multi-objective processing and real-time control decision-making. These shortcomings can lead to suboptimal decisions, potentially creating dangers in extreme situations. The limitations of real-time control decision-making manifest in the fact that rule-driven systems struggle to cope with unforeseen scenarios. Furthermore, the use of complex algorithms can exceed the computing power of the onboard ECU, resulting in response delays, impacting control effectiveness, and posing risks. Furthermore, conventional methods adjust multiple weights simultaneously in each generation, causing drastic changes in the search direction and compromising algorithm stability.

[0005] The steady-state weight adaptive algorithm, however, can decompose a target into multiple objectives during multi-objective conflict optimization, adjusting weights based on real-time operating conditions to prioritize key objectives. This weight adaptive mechanism enables the algorithm to quickly adapt to parameter changes, providing robust dynamic response capabilities. The steady-state weight adaptive algorithm can effectively address the shortcomings of existing stability control algorithms in multi-objective processing and real-time control decision-making. However, when controlling vehicle stability, the steady-state weight adaptive algorithm may experience weight fluctuations due to improper learning rate settings or imperfect feedback mechanisms, leading to fluctuations in control commands. It can also be difficult to make differentiated adjustments for different road conditions, resulting in poor scenario adaptability. Summary of the Invention

[0006] In order to solve the above problems, the present invention proposes a vehicle body stability control method, system, equipment, medium and product. The present invention realizes vehicle body stability control, has the advantages of high efficiency of multi-objective optimization, high adaptability to dynamic working conditions and reduced parameter sensitivity, can meet the needs of online multi-objective optimization, and has broad application prospects.

[0007] According to some embodiments, a first solution of the present invention provides a vehicle body stability control method, which adopts the following technical solution:

[0008] A vehicle body stability control method, comprising:

[0009] Real-time acquisition of vehicle status information and pre-processing to obtain yaw rate, sideslip angle, road adhesion coefficient, steering wheel angle and slope;

[0010] The problem is modeled using steering wheel angle, slope, and road adhesion coefficient as constraints, the stability control objective is decomposed into three objectives, and the initial population of solutions and the initial weight set of all objectives are generated;

[0011] If the rate of change of the objective function of all targets exceeds the set threshold, it is determined that the environment has undergone a sudden change; when the rate of change of the objective function is less than the set threshold, the vehicle enters a steady state, and the weight vector is adjusted according to the current working conditions;

[0012] Solving the objective function based on the adjusted weight vector, generating a braking force distribution instruction and a driving torque instruction, and controlling the vehicle electronic stability control system based on the braking force distribution instruction and the driving torque optimization instruction;

[0013] Receive vehicle status feedback in real time to determine whether the environment has undergone a sudden change. If not, return to the weight adjustment process when the scheduled monitoring time is reached; if so, re-trigger the weight adjustment process.

[0014] According to some embodiments, a second solution of the present invention provides a vehicle body stability control system, which adopts the following technical solution:

[0015] A vehicle body stability control system comprising:

[0016] The data acquisition and preprocessing module is configured to obtain vehicle status information in real time and perform preprocessing to obtain yaw rate, sideslip angle, road adhesion coefficient, steering wheel angle and slope;

[0017] The problem modeling and initialization module is configured to model the problem using steering wheel angle, slope, and road adhesion coefficient as constraints, decompose the stability control objective into three objectives, and generate an initial population of solutions and an initial set of weights for all objectives;

[0018] The steady-state detection and weight adjustment module is configured to determine that the environment has undergone a sudden change if the rate of change of the objective function of all targets exceeds a set threshold. If the rate of change of the objective function is less than the set threshold, the vehicle enters a steady state and the weight vector is adjusted according to the current operating conditions.

[0019] an output optimization control instruction module, configured to solve an objective function based on the adjusted weight vector, generate a braking force distribution instruction and a driving torque instruction, and control the vehicle electronic stability control system based on the braking force distribution instruction and the driving torque optimization instruction;

[0020] The closed-loop feedback readjustment module is configured to receive vehicle status feedback in real time and determine whether there is a sudden change in the environment. If not, it returns to the weight adjustment process when the scheduled monitoring time is reached; if so, it re-triggers the weight adjustment process.

[0021] According to some embodiments, a third aspect of the present invention provides a computer-readable storage medium.

[0022] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of a vehicle body stability control method as described in the first aspect above.

[0023] According to some embodiments, a fourth aspect of the present invention provides a computer device.

[0024] A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, the steps of the vehicle body stability control method described in the first aspect are implemented.

[0025] According to a fifth aspect of the present invention, there is provided a computer program product or computer program according to some embodiments.

[0026] The present invention provides a computer program product or computer program, comprising 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 executes the computer instructions, causing the computer device to perform the steps of the vehicle body stability control method described in the first aspect above.

[0027] Compared with the prior art, the present invention has the following beneficial effects:

[0028] This paper proposes a vehicle stability control method based on a steady-state weighted adaptive algorithm, aiming to address the shortcomings of traditional vehicle stability control algorithms in multi-objective processing and real-time control decision-making. This method uses sensors to acquire real-time vehicle state data, including yaw rate, sideslip angle, wheel speed, steering wheel angle, lateral and longitudinal acceleration, driving torque, and braking force. It then optimizes vehicle stability control using a steady-state weighted adaptive algorithm, dynamically adjusting the weights of each control objective to achieve multi-objective optimization and real-time control decision-making. By decomposing the global optimization problem into multiple objectives and solving them in parallel using an onboard multi-core processor, computational latency is reduced, improving real-time performance and computational efficiency. Furthermore, by introducing a steady-state trigger mechanism that triggers weight updates only when the rate of change of the system state falls below a threshold, redundant computation is reduced, further improving real-time performance and computational efficiency. This method effectively addresses the shortcomings of vehicle stability control algorithms in multi-objective processing and real-time control decision-making.

[0029] The present invention proposes a new method for optimizing the vehicle body stability control algorithm based on a steady-state weight adaptive algorithm. This method resolves multi-objective dynamic conflicts and improves control flexibility. It can automatically adjust the weights of each objective based on the real-time vehicle state and external environment, balancing comfort and safety in daily driving while enhancing robustness under extreme conditions. At the same time, it can approximate the global optimal solution under different operating conditions by decomposing multiple objectives and adaptively assigning weights, rather than compromising on a single objective, thus avoiding falling into a local optimal situation. By decomposing the global optimization problem into multiple objectives, the on-board multi-core processor is used to solve them in parallel, shortening the calculation. A steady-state trigger mechanism is introduced, which triggers weight updates only when the system state change rate is lower than a threshold, reducing redundant calculations and delays, thereby improving real-time performance and computational efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] The accompanying drawings, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0031] Figure 1 This is a control concept diagram of a vehicle body stability control method according to an embodiment of the present invention;

[0032] Figure 2 is a processing flow chart of the main loop in an embodiment of the present invention;

[0033] Figure 3 Schematic diagram of sensors and parameter collection in an embodiment of the present invention;

[0034] Figure 4 Schematic diagram of problem modeling and initialization in an embodiment of the present invention;

[0035] Figure 5This is a schematic diagram of the archive maintenance process in an embodiment of the present invention;

[0036] Figure 6 Schematic diagram of adjusting population structure (adding weights) in an embodiment of the present invention;

[0037] Figure 7 Schematic diagram of adjusting population structure (deleting weights) in an embodiment of the present invention. DETAILED DESCRIPTION

[0038] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0039] It should be noted that the following detailed descriptions are illustrative and intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which the present invention belongs.

[0040] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is intended to include the plural form. In addition, it should be understood that when the terms "comprise" and / or "include" are used in this specification, they indicate the presence of features, steps, operations, devices, components and / or combinations thereof.

[0041] In the absence of conflict, the embodiments of the present invention and the features thereof may be combined with each other.

[0042] The present invention is mainly used for optimizing vehicle body stability control algorithms, mainly detecting the vehicle state through sensors, and then dynamically adjusting the vehicle by controlling the brakes and engine output. However, traditional vehicle body stability control algorithms need to complete sensor data processing, state estimation, and control decisions in a very short time, which places high demands on the algorithm's computational efficiency and requires the realization of multiple control objectives. It is difficult for the algorithm to achieve the most balanced solution in all scenarios, and the overall performance is limited. The present invention proposes a steady-state weight adaptive algorithm based on decomposition, which obtains data such as yaw rate, center of mass sideslip angle, wheel speed, steering wheel angle, lateral and longitudinal acceleration, driving torque, and braking force through sensors, and performs preprocessing. The control target is then decomposed into multiple sub-problems, and steady-state control and weight adjustment are performed. The sub-problems are solved based on the updated weights, and optimization instructions are output. Finally, closed-loop feedback is performed for re-adjustment. This method uses a steady-state weight adaptive algorithm to dynamically adjust weights based on the vehicle's real-time state and external environment. By breaking down multiple objectives into subproblems and adaptively assigning weights, it leverages the vehicle's multi-core processors for parallel solutions, shortening computational latency. A flexible weight removal mechanism gradually adjusts weights, avoiding the impact of drastic fluctuations during weight adjustment and maintaining population diversity. By introducing a steady-state weight adaptive algorithm, the electronic stability system (ESP) can be upgraded from a "rule-driven" approach to a "goal-driven" one, significantly improving vehicle safety, controllability, and user experience in complex scenarios.

[0043] Example 1

[0044] This embodiment provides a vehicle body stability control method, including:

[0045] Real-time acquisition of vehicle status information and pre-processing to obtain yaw rate, center of mass sideslip angle, road adhesion coefficient, steering wheel angle and slope;

[0046] The problem is modeled using steering wheel angle, slope, and road adhesion coefficient as constraints, the stability control objective is decomposed into three objectives, and the initial population of solutions and the initial weight set of all objectives are generated;

[0047] If the rate of change of the objective function of all targets exceeds the set threshold, it is determined that the environment has undergone a sudden change; when the rate of change of the objective function is less than the set threshold, the vehicle enters a steady state, and the weight vector is adjusted according to the current working conditions;

[0048] Solving the objective function based on the adjusted weight vector, generating a braking force distribution instruction and a driving torque instruction, and controlling the vehicle electronic stability control system based on the braking force distribution instruction and the driving torque optimization instruction;

[0049] Receive vehicle status feedback in real time to determine whether the environment has undergone a sudden change. If not, return to the weight adjustment process when the scheduled monitoring time is reached; if so, re-trigger the weight adjustment process.

[0050] like Figure 1 As shown, the specific process of this embodiment includes:

[0051] First, the algorithm is initialized through two parts: sensor and parameter acquisition and problem modeling and initialization;

[0052] Environmental changes are detected based on the rate of change of the objective function. If the rate of change of the objective functions of all sub-goals exceeds the set threshold, it is determined that the environment has undergone a sudden change. When the rate of change of the objective function is less than the set threshold, the vehicle enters a steady state, and the weight vector is adjusted according to the current working conditions:

[0053] Based on the sensor data of the current working condition, determine whether there is a corresponding scene feature in file A; if so, use the weight vector corresponding to the current scene feature to obtain an adjusted weight vector; the sensor data includes yaw rate, center of mass slip angle, road adhesion coefficient, steering wheel angle, and slope; the scene features include yaw rate error, center of mass slip angle, and brake pressure fluctuation;

[0054] If not, enter the main loop and determine whether the termination condition is met. The termination condition includes several parts. If one of them is met, the current loop will be terminated, including:

[0055] (1) Archiving for N consecutive generations (e.g., N=3) No change, indicating that The solution in is sufficient to meet the current situation, and the optimization output is Extracted from

[0056] (2) Reaching the maximum number of iterations , the maximum number of iterations is the upper limit of the preset algorithm running time (such as = 50 iterations), ensuring that the control cycle is ≤ 10ms (corresponding to a 100Hz real-time control frequency), and that word iteration times out and is terminated forcibly;

[0057] (3) The vehicle state is detected to have entered a safe range, the yaw rate error is less than a certain range (e.g., 2°), and the sideslip angle is less than a certain range (e.g., 3°);

[0058] (4) The ESP / ABS system actively intervenes, immediately terminates optimization and transfers control.

[0059] If the termination condition is not met, continue to iterate the main loop, enter the offspring generation and environment selection part, update the population and perform archive maintenance, use the archive to generate a candidate solution set, and then determine whether the current main loop has entered the inner loop;

[0060] If not, then enter the inner loop:

[0061] Use the candidate solution set to adjust the initial population and weight set until the number of inner loop iterations reaches 90% of the maximum number of iterations, then output the adjusted population and weight set;

[0062] Inner loop determines the number of iterations Whether the maximum number of iterations has been reached The previous iterations mainly explored the solution space extensively and dynamically adjusted the weights to adapt to different driving scenarios.

[0063] The final 10% of frozen weight adjustment focuses on convergence optimization and optimizes the optimized strategy. Mapped to the ESP actuator, the effect is verified through a closed-loop sensor feedback loop, triggering a new round of optimization.

[0064] If it is reached, determine whether the termination condition is met. If not, continue to iterate the main loop; when the stop condition is met, output the adjusted weight vector.

[0065] Based on the output results, real-time control execution and control parameter mapping are performed; the ESP (Electronic Stability Control) actuator is activated, the vehicle status is fed back, and the system enters the observation state. If the environment suddenly changes again, or if the preset detection time (e.g., 1 second) is reached, the system again checks whether there are corresponding scene features in archive A. If not, the system enters the main loop.

[0066] The method includes five steps: sensor parameter acquisition and preprocessing, problem modeling and initialization, steady-state detection and weight adjustment, output optimization control instructions, closed-loop feedback and readjustment, as follows:

[0067] Step 1: Sensor parameter acquisition and preprocessing.

[0068] Step 1.1: The vehicle's yaw rate, sideslip angle, wheel speed, steering wheel angle, lateral and longitudinal acceleration, driving torque and braking force, road adhesion coefficient and slope data are all obtained through sensors or calculated.

[0069] like Figure 2As shown, the sensors used in this embodiment include wheel speed sensors, a steering wheel angle sensor, a torque sensor, an image sensor, and an IMU (Inertial Measurement Unit) sensor. The wheel speed sensors are installed on all four wheels, capturing data from the four wheel speed sensors for calculating slip ratio and inter-wheel speed. The wheel speed sensors acquire the speed of each wheel and combine this with the actual vehicle speed acquired by the IMU sensor to calculate the vehicle slip ratio. The steering wheel angle sensor is installed in the steering column below the steering wheel to acquire the steering wheel angle. A driving torque sensor is installed near the drive shaft, and a braking torque sensor is installed at the brake caliper mounting point to obtain driving torque and braking torque. Image sensors and infrared sensors are used to identify road surface conditions and estimate road adhesion coefficient and slope. The IMU sensor is installed at the bottom of the vehicle cockpit and transmits vehicle posture data to the controller via CAN communication signals. It can acquire lateral and longitudinal acceleration, actual vehicle speed, and yaw rate. Combined with the wheel speed sensors, the sideslip angle is calculated according to the following formula.

[0070] First, the longitudinal speed is estimated by the wheel speed sensor .

[0071] Then the yaw rate measured by the IMU gyroscope and IMU accelerometer measurements Points .

[0072] (1);

[0073] (2);

[0074] Where, is the longitudinal velocity, is the lateral velocity, is the yaw angular velocity, is the sideslip angle of the center of mass.

[0075] Step 1.2: Apply Kalman filtering to the acquired data to eliminate noise.

[0076] Step 2: Problem modeling and initialization, such as Figure 3 As shown in the figure, first, the yaw rate, wheel speed, steering wheel angle, lateral and longitudinal acceleration, driving torque and braking force data are obtained through sensors; and the center of mass slip angle, road adhesion coefficient and slope, and slip rate data are estimated. Then, the yaw rate, center of mass slip angle, and road adhesion coefficient are used as inputs, and the road adhesion coefficient, slope, and steering wheel angle are used as constraints. The control target is decomposed into multiple targets, and then the population is initialized. and weight set .

[0077] Step 2.1: Determine the decision variables: four-wheel braking force distribution;

[0078] Step 2.2: Determine the input parameters: yaw rate ( )、center of mass side slip angle( ), road adhesion coefficient ( );

[0079] Step 2.3: Construct the objective function (decompose the control objective into multiple optimization objectives):

[0080] ①Goal 1: This goal is mainly to minimize the yaw rate error. By optimizing this goal, the stability of the vehicle can be improved.

[0081] (3);

[0082] Where, is the objective function of goal 1, is the required yaw rate, is the actual yaw angular velocity.

[0083] ②Goal 2: This goal is mainly to suppress the sideslip angle of the center of mass. By optimizing this goal, the safety of the vehicle can be improved.

[0084] (4);

[0085] Where, is the objective function of goal 2, is the sideslip angle of the center of mass.

[0086] ③Goal 3: This goal is mainly to reduce brake pressure fluctuations. By optimizing this goal, the comfort of the vehicle can be improved.

[0087] (5);

[0088] Where, is the objective function of objective 3, For pressure fluctuations.

[0089] Step 2.4: Determine the constraints: road adhesion coefficient, slope, steering wheel angle and other dynamic limitations.

[0090] Step 2.5: Initialize the population and weight set , specifically:

[0091] ① Randomly generate an initial solution set, which is the population , each solution represents a set of torque allocation strategies.

[0092] ② Initialize the weight set The uniformly distributed three-dimensional vector is initially set to [0.5, 0.3, 0.2], with a sum of 1, corresponding to the weights of the three targets. Middle weight vector Contains a set of weight data, that is, ;in, Indicates the Solution On target The corresponding weight, .

[0093] The stability control target is decomposed into the stability weight of target 1, the safety weight of target 2 and the comfort weight of target 3; target 1 corresponds to the stability weight, that is, the yaw rate error weight is , Target 2 corresponds to the safety weight, that is, the center of mass side slip angle suppression weight is , Target 3 corresponds to the comfort weight, that is, the brake judder suppression weight is .Right now .

[0094] Step 3: Steady-state detection and weight adjustment, and enter the main loop, such as Figure 4 shown.

[0095] Step 3.1: Determine the set of non-dominated solutions, specifically:

[0096] Dominant solution definition: solution Outperforms the solution on at least one objective , and is not inferior to the solution in other objectives ,but Dominate .

[0097] Definition of non-dominated solution: If the solution is not dominated by any other solution, then It is a non-dominated solution.

[0098] First, calculate the yaw rate error, center of mass sideslip angle, and brake pressure fluctuation of all solutions. Filter the non-dominated solutions through the Pareto dominance relationship and save them in the archive. , that is, archive It contains all non-dominated solutions and their corresponding scene features and weight vectors, as well as the mapping relationship between scene features, non-dominated solutions and weight vectors.

[0099] 2) Neighborhood Update

[0100] Archive Each weight vector in is updated in the neighborhood, where

[0101] Neighborhood definition: For each weight vector ,calculate The Euclidean distance to other weight vectors, select the nearest as the weight vector neighborhood.

[0102] parameter : Set to 10-20% of the population size to balance local search and diversity.

[0103] Step 3.2: Steady-state detection (combined with real-time data), specifically:

[0104] ①Calculate the rate of change of the objective function:

[0105] | (6);

[0106] Where, for The rate of change of the objective function at time for Time target The objective function, for Time target The objective function is The rate of change of the objective function exceeds the rate of change threshold ,Right now , then it is judged that the environment has undergone a sudden change;

[0107] ②If If it is considered to have entered a steady state, the weights will be adjusted according to the current working conditions and the process will proceed to step 3.3.

[0108] In addition, in response to environmental mutations in special scenarios, the weights are adjusted in the following ways:

[0109] Low adhesion road surface (when the road adhesion coefficient is lower than a certain range or the slip rate of one wheel is significantly higher than that of other wheels): Increase the safety weight , reduce the comfort weight .

[0110] The slip rate calculation formula is as follows:

[0111] (7);

[0112] Where, For wheels The slip rate, For wheels Linear speed, The actual speed of the vehicle measured by the IMU sensor.

[0113] wheel The linear velocity calculation formula is as follows:

[0114] (8);

[0115] Where, For wheels Linear speed, is the wheel speed measured by the wheel speed sensor, The effective radius of the tire, which needs to be calibrated regularly to compensate for the effects of tire pressure or wear.

[0116] Aggressive driving situations (large changes in lateral speed and acceleration): Increase stability weighting , reduce the comfort weight .

[0117] Step 3.3: Based on the sensor data of the current working condition, determine whether there is a corresponding scene feature in archive A. If not, enter the main loop:

[0118] Step 3.3.1: Offspring generation and environmental selection, specifically:

[0119] Step 3.3.1.1: Offspring generation:

[0120] Initialize the descendants collection , that is, the descendant set Initialized to empty, used to store newly generated individuals;

[0121] For each parent individual , perform the following operations: Based on the parent individual and populations Generate an offspring ,Will join in .

[0122] Step 3.3.1.2: Environment selection: first merge the initial population With descendant collection Get the merged collection and then perform environment screening as follows:

[0123] 1) Decomposition method (MOEA / D): For each weight vector corresponding to each solution in the combined set , compute the weighted sum of solutions or Tchebycheff values ​​(zeros of the Chebycheff polynomials), retaining the optimal solution for each neighborhood of each weight vector.

[0124] Normalize the solution of each target to obtain the normalized solution of each target;

[0125] In order to ensure that the values ​​of different objective functions are compared on the same scale, the steady-state weight adaptive algorithm normalizes the objective function values. The normalization formula is as follows:

[0126] (9);

[0127] Where, Is the solution In the Normalized value on the target. and They are all the solutions in the archive in The minimum and maximum values ​​on the targets.

[0128] 2) Non-dominated sorting: Assist in eliminating the dominated solutions in the normalized solution of each target to obtain the updated population , ensuring convergence and obtaining the corresponding updated weight set .

[0129] Step 3.3.2: Archive maintenance, such as Figure 5 As shown;

[0130] Step 3.3.2.1: Non-dominated selection, set the offspring The non-dominated solutions in are added to the archive , remove the old solution dominated by the new solution and get the updated archive ;

[0131] Step 3.3.2.2: Normalize the updated archive The individuals in the target space are normalized to obtain the normalized archive .

[0132] Step 3.3.2.3: If normalized archive The size exceeds the preset capacity , through crowding distance pruning, we get the candidate solution set , to ensure the diversity and distribution of archives, the specific process is as follows:

[0133] if Than the preset capacity Large, based on normalized archive Screen and get the parent individual set ;

[0134] Archive from normalization via non-dominated sorting Remove parent individuals , and the offspring elite set is ;

[0135] Traverse the elite set of descendants Each solution in , one by one Add parent individual set →Intermediate candidate solution set ;

[0136] In the middle candidate solution set In the , choose to make the crowding function The largest solution and removed;

[0137] Until all solutions in the offspring elite set are traversed, the candidate solution set is obtained .

[0138] Archives after this Replaced by candidate solution set , ensuring the diversity and distribution of archives.

[0139] The congestion calculation formula is as follows:

[0140] (10);

[0141] Where, For the solution In the solution set The congestion in .

[0142] Is the solution reconciliation The relative distance between them is calculated as follows:

[0143] (11);

[0144] Where, Is the solution reconciliation The Euclidean distance between is the neighborhood radius, which is the same size as the previous parameter related.

[0145] Step 3.3.3: Determine whether this main loop has entered the inner loop. If not, enter the inner loop, that is, go to step 3.3.4; if yes, go directly to step 4.

[0146] Step 3.3.4: Inner loop, specifically including:

[0147] Step 3.3.4.1: Adjust the population structure (add weights), such as Figure 6 As shown, based on the candidate solution set, high-quality solutions are selected and added to the updated population, and the weight vectors corresponding to the high-quality solutions are added to the updated weight set to obtain the adjusted population and the adjusted weight set, which are specifically:

[0148] First create an empty elite solution set , and initialize it.

[0149] Traverse the candidate solution set Each solution in , and for each solution Generate a weight vector ,in, It is Solution On target The corresponding weight, .

[0150] Define solution Neighborhood index set .

[0151] like In the weight vector If it is better than all neighboring solutions, it will be added to the elite solution set .

[0152] In this embodiment, the scaling function From the candidate solution set The solution with good convergence is selected, and then the solution that can best improve the diversity of the population is identified to promote the uniform distribution of the solution set in the target space.

[0153] Scaling function: Mathematically, let and For solution, For solution On target The corresponding weights are defined. In weight Better than ,Right now ,but

[0154] (12);

[0155] If the solution In weight If it is better than all neighboring solutions, it can be added to the elite solution set .

[0156] Until the traversal is completed, calculate the elite solution set Each solution To the original population The minimum distance to other solutions in ;

[0157] Select the solution with the smallest distance and the longest distance (i.e. The largest corresponding element ) . Will solve Join the update population , the corresponding maximum weight vector Add an updated weight set and finally return the adjusted population and adjusted weight set to ensure that the elite solution selection is based on a high-quality solution set accumulated over a long period of optimization.

[0158] Computational elite solution set The diversity contribution of each solution in (minimum distance from the population) is calculated and the solution that promotes diversity the most is selected.

[0159] The minimum distance from the population is calculated as follows:

[0160] (13);

[0161] The minimum distance is :

[0162] (14);

[0163] Then among all the calculated distances, find the minimum value as To the population The minimum distance.

[0164] Step 3.3.4.2: Adjust the population structure (delete weights), such as Figure 7 As shown, the solution with the most repetitions in the candidate solution set is selected for deletion, and the corresponding weight vector is deleted from the adjusted weight set to obtain a new generation of weight set, which is:

[0165] Determine the candidate solution set Is there a repeated solution in ?

[0166] If there are repeated solutions, find the solution with the largest number of repetitions to generate a repeated solution set. Find the solution with the largest scaling function in the repeated solution set and delete it from the candidate solution set. At the same time, delete the corresponding weight vector in the adjusted weight set to obtain a new generation weight set.

[0167] If there are no duplicate solutions, the congestion of each solution is calculated, and the solution with the largest congestion is found and deleted from the candidate solution set. At the same time, the corresponding weight vector in the adjusted weight set is deleted to obtain a new generation of weight set.

[0168] In the candidate solution set Among duplicate solutions, solutions with poor convergence are prioritized for deletion. If there are no duplicates, the solution with the smallest contribution to population diversity is deleted to ensure a uniform population distribution. The core goal is to balance the diversity and quality of solutions by dynamically adjusting the population structure. From the set of solutions with the most duplicates, the solution with the highest objective function value is selected for deletion. The core goal is to balance the diversity and quality of solutions by dynamically adjusting the population structure.

[0169] The operation object for removing duplicate solutions and low diversity solutions is the candidate solution set This design is done by dynamic cleaning The low-quality solution in the population ensures that it will be merged into the original population later. The candidate solutions in are both high-quality and diverse, thus achieving a balance between convergence and distribution at the global level.

[0170] The direct operation object for adjusting the population structure is the candidate solution set , but indirectly affects the original population through :

[0171] Elite solution injection: from the candidate solution set Select high-quality solutions to join the update population to improve the quality of the update population.

[0172] Eliminating inferior solutions: purifying the candidate solution set to avoid low-quality solutions from entering the updated population of subsequent iterations.

[0173] Step 3.3.4.3: Neighborhood update: Update the neighborhood of each weight vector in the new generation weight set, and adjust the control by weight vector The direction of generating new solutions.

[0174] Step 3.3.4.4: Loop through the inner loop until the inner loop iteration count Reached the maximum number of iterations If the number of users is 90%, go to step 4.

[0175] Step 4: Output optimization control instructions:

[0176] 1) Real-time control execution, ensuring that the control instructions generated by the algorithm complete the entire chain from calculation to execution in a very short time (e.g., ≤10ms);

[0177] The control parameter mapping solves the target based on the updated weights and converts it into specific physical instructions that the ESP actuator can understand, generating braking force distribution and driving torque instructions.

[0178] 2) ESP actuator action, sending braking force distribution and driving torque instructions to the brake actuator and motor controller in the ESP system, physically changing the vehicle dynamic state through the ESP system (such as the brake hydraulic unit and electronic stability program), and the brake actuator and motor controller in the ESP system perform corresponding actions.

[0179] Step 5: Closed-loop feedback and readjustment:

[0180] Receive vehicle status feedback and monitor control effectiveness in real time through the sensor network, forming a closed-loop verification. Continuously monitor the optimization results returned previously, and repeat the process of step 3.

[0181] Based on the rate of change of the objective function, determine whether the environment has undergone a sudden change. If not, return to step 3 when the scheduled monitoring time is reached;

[0182] If yes, it will trigger step 3 again to adjust the weight.

[0183] The above process is repeated cyclically to determine the optimal weight vector for vehicle stability control.

[0184] Example 2

[0185] This embodiment provides a vehicle body stability control system, including:

[0186] The data acquisition and preprocessing module is configured to obtain vehicle status information in real time and perform preprocessing to obtain yaw rate, sideslip angle, road adhesion coefficient, steering wheel angle and slope;

[0187] The problem modeling and initialization module is configured to model the problem using steering wheel angle, slope, and road adhesion coefficient as constraints, decompose the stability control objective into three objectives, and generate an initial population of solutions and an initial set of weights for all objectives;

[0188] The steady-state detection and weight adjustment module is configured to determine that the environment has undergone a sudden change if the rate of change of the objective function of all targets exceeds a set threshold. If the rate of change of the objective function is less than the set threshold, the vehicle enters a steady state and the weight vector is adjusted according to the current operating conditions.

[0189] an output optimization control instruction module, configured to solve an objective function based on the adjusted weight vector, generate a braking force distribution instruction and a driving torque instruction, and control the vehicle electronic stability control system based on the braking force distribution instruction and the driving torque optimization instruction;

[0190] The closed-loop feedback readjustment module is configured to receive vehicle status feedback in real time and determine whether there is a sudden change in the environment. If not, it will return to the weight adjustment process when the scheduled monitoring time is reached; if so, it will re-trigger the weight adjustment process.

[0191] The examples and application scenarios implemented by the above modules and corresponding steps are the same, but are not limited to the contents disclosed in the above embodiment 1. It should be noted that the above modules as part of the system can be executed in a computer system such as a set of computer executable instructions.

[0192] The descriptions of the various embodiments in the above embodiments have different focuses. For parts not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0193] The proposed system can be implemented in other ways. For example, the system embodiment described above is merely illustrative. For example, the above module division is only a logical function division. In actual implementation, other division methods may be used. For example, multiple modules can be combined or integrated into another system, or some features can be ignored or not implemented.

[0194] Example 3

[0195] This embodiment provides a computer-readable storage medium having a computer program stored thereon. When the program is executed by a processor, the steps of the vehicle body stability control method described in the first embodiment are implemented.

[0196] Example 4

[0197] This embodiment provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, the steps of the vehicle body stability control method described in the first embodiment are implemented.

[0198] Example 5

[0199] This embodiment provides a computer program product or 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 executes the computer instructions, causing the computer device to perform the steps of the vehicle body stability control method described in the first embodiment.

[0200] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of hardware embodiments, software embodiments, or embodiments combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage and optical storage, etc.) containing computer-usable program code.

[0201] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0202] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0203] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0204] Those skilled in the art will appreciate that all or part of the processes in the above-described method embodiments can be implemented by instructing the relevant hardware through a computer program. The program can be stored in a computer-readable storage medium, and when executed, the program can include the processes in the above-described method embodiments. The storage medium can be a magnetic disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM).

[0205] Although the above describes the specific embodiments of the present invention in conjunction with the accompanying drawings, it is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art on the basis of the technical solution of the present invention without any creative work are still within the scope of protection of the present invention.

Claims

1. A vehicle body stability control method, characterized in that: include: Real-time acquisition of vehicle status information and pre-processing to obtain yaw rate, center of mass sideslip angle, road adhesion coefficient, steering wheel angle and slope; The problem is modeled using steering wheel angle, slope, and road adhesion coefficient as constraints, the stability control objective is decomposed into three objectives, and the initial population of solutions and the initial weight set of all objectives are generated; If the rate of change of the objective function of all targets exceeds the set threshold, it is determined that the environment has undergone a sudden change; when the rate of change of the objective function is less than the set threshold, the vehicle enters a steady state, and the weight vector is adjusted according to the current working conditions, specifically: Determine non-dominated solutions based on the initial population and the initial weight set, and save all non-dominated solutions and their corresponding scene features and weight vectors, as well as the mapping relationship between scene features, non-dominated solutions and weight vectors into an archive; Based on the sensor data of the current working condition, determine whether there is a corresponding scene feature in the archive; If yes, the adjusted weight vector is obtained by using the weight vector corresponding to the current scene feature; If not, enter the main loop and use the archive to generate candidate solution sets until the termination condition is met and the adjusted weight vector is output; During each iteration of the main loop, it is necessary to determine whether the current main loop has entered the inner loop; If not, enter the inner loop and use the candidate solution set to adjust the initial population and weight set until the number of inner loop iterations reaches 90% of the maximum number of iterations, and then output the adjusted population and weight set; If yes, then determine whether the termination condition is met. If not, continue to iterate the main loop. If so, output the adjusted weight vector. Solving the objective function based on the adjusted weight vector, generating a braking force distribution instruction and a driving torque instruction, and controlling the vehicle electronic stability control system based on the braking force distribution instruction and the driving torque optimization instruction; Receive vehicle status feedback in real time to determine whether the environment has undergone a sudden change. If not, return to the weight adjustment process when the scheduled monitoring time is reached; if so, re-trigger the weight adjustment process.

2. A vehicle body stability control method as claimed in claim 1, characterized in that: The stability control objective is decomposed into three objectives, generating an initial population of solutions and an initial weight set of all objectives, including: Decompose the stability control objective into the stability weight of objective 1, the safety weight of objective 2, and the comfort weight of objective 3, and the sum of the three weights is equal to 1; Among them, the objective function of goal 1 is to minimize the yaw rate error, the objective function of goal 2 is to minimize the sideslip angle of the center of mass, and the objective function of goal 3 is to minimize the brake pressure fluctuation; Solve the three objectives and obtain the initial population and the initial weight set of all objectives.

3. The vehicle body stability control method according to claim 1, characterized in that: The main loop is entered, and the candidate solution set is generated by using the archive until the termination condition is met and the adjusted weight vector is output, specifically: Take each solution in the initial population as a parent individual and generate a set of offspring based on the parent individuals and the initial population; After merging the initial population and the offspring set, environmental screening is performed to obtain the updated population and the updated weight set; Add the non-dominated solutions in the offspring set to the archive and remove the old solutions dominated by the new solutions to obtain the updated archive; Normalize the target space of the individuals in the updated archive to obtain a normalized archive; If the normalized archive size exceeds the preset capacity, the candidate solution set is obtained by pruning through the crowding distance; The main loop is iteratively repeated until the termination condition is met and the adjusted weight vector is output.

4. A vehicle body stability control method as claimed in claim 1, characterized in that: If the normalized archive size exceeds the preset capacity, the candidate solution set is obtained by pruning the congestion distance, specifically: Screening is performed based on the normalized archive to obtain the parent individual set; By removing the parent individuals from the normalized archive through non-dominated sorting, the elite set of offspring is obtained as follows; Traverse each solution in the offspring elite set and add them one by one to the parent individual set to generate an intermediate candidate solution set; In the set of intermediate candidate solutions, the solution that maximizes the congestion function is selected and removed; Until all solutions in the offspring elite set are traversed, the candidate solution set is obtained.

5. The vehicle body stability control method according to claim 1, characterized in that: The inner loop uses the candidate solution set to adjust the updated population and update the weight set, specifically: Based on the candidate solution set, high-quality solutions are selected and added to the updated population, and the weight vectors corresponding to the high-quality solutions are added to the updated weight set to obtain the adjusted population and the adjusted weight set; Select the solution with the most repetitions in the candidate solution set and delete it, and delete the corresponding weight vector from the adjusted weight set to obtain a new generation weight set; Perform neighborhood updates on each weight vector in the new generation weight set; The inner loop is iterated based on the above process until the number of inner loop iterations reaches 90% of the maximum number of iterations.

6. A vehicle body stability control system, characterized in that: include: The data acquisition and preprocessing module is configured to obtain vehicle status information in real time and perform preprocessing to obtain yaw rate, sideslip angle, road adhesion coefficient, steering wheel angle and slope; The problem modeling and initialization module is configured to model the problem using steering wheel angle, slope, and road adhesion coefficient as constraints, decompose the stability control objective into three objectives, and generate an initial population of solutions and an initial set of weights for all objectives; The steady-state detection and weight adjustment module is configured to determine that the environment has undergone a sudden change if the rate of change of the objective function of all targets exceeds a set threshold. When the rate of change of the objective function is less than the set threshold, the vehicle enters a steady state and the weight vector is adjusted according to the current operating conditions. Specifically, Determine non-dominated solutions based on the initial population and the initial weight set, and save all non-dominated solutions and their corresponding scene features and weight vectors, as well as the mapping relationship between scene features, non-dominated solutions and weight vectors into an archive; Based on the sensor data of the current working condition, determine whether there is a corresponding scene feature in the archive; If yes, the adjusted weight vector is obtained by using the weight vector corresponding to the current scene feature; If not, enter the main loop and use the archive to generate candidate solution sets until the termination condition is met and the adjusted weight vector is output; During each iteration of the main loop, it is necessary to determine whether the current main loop has entered the inner loop; If not, enter the inner loop and use the candidate solution set to adjust the initial population and weight set until the number of inner loop iterations reaches 90% of the maximum number of iterations, and then output the adjusted population and weight set; If yes, then determine whether the termination condition is met. If not, continue to iterate the main loop. If so, output the adjusted weight vector. an output optimization control instruction module, configured to solve an objective function based on the adjusted weight vector, generate a braking force distribution instruction and a driving torque instruction, and control the vehicle electronic stability control system based on the braking force distribution instruction and the driving torque optimization instruction; The closed-loop feedback readjustment module is configured to receive vehicle status feedback in real time and determine whether there is a sudden change in the environment. If not, it returns to the weight adjustment process when the scheduled monitoring time is reached; if so, it re-triggers the weight adjustment process.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of the vehicle body stability control method according to any one of claims 1 to 5 are implemented.

8. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the steps of the vehicle body stability control method according to any one of claims 1 to 5 are implemented.

9. A computer program product, characterized in that The computer program product includes a computer program, and when the computer program is executed by a processor, the computer program implements the steps in the vehicle body stability control method according to any one of claims 1 to 5.

Images