Automatic power-assisted steering system control method

Abstract

The invention provides an automatic power-assisted steering system control method, which belongs to the field of PID control optimization and specifically comprises the following steps: S1, acquiring current steering angle data of a vehicle in real time through a steering angle sensor; s2, calculating a target power-assisted steering torque based on the steering angle data and the current dynamic parameters of the vehicle; s3, the torque PID controller generates a corresponding control signal according to the error between the target power-assisted steering torque and the actual power-assisted steering torque; s4, control parameters of the torque PID controller are optimized through an improved language education algorithm; the improved language education algorithm is specifically realized as follows: D1, using a path backtracking feedback mechanism; d2, using a dynamic target guidance and gravitational effect optimization strategy; and S5, the control signal ensures that the output torque of the power-assisted motor is highly matched with the steering requirement of the vehicle by precisely adjusting the output torque of the power-assisted motor, and therefore precise control over automatic power-assisted steering of the vehicle is achieved.

Description

Technical Field

[0001] The present invention belongs to the field of PID control optimization, and particularly relates to a control method for an electric power steering system. Background Art

[0002] An electric power steering system is an important steering assistance system in modern automobiles, widely used to improve driving comfort and maneuverability; the traditional mechanical steering system transmits the driver's operating force to the wheels through a steering column and a gear device, and the driver needs to apply a large force to complete steering, especially at low speeds or when parking, the steering operation becomes more strenuous; with the development of technology, the electric power steering system uses an electric motor instead of the traditional mechanical device, and adjusts the assist output of the electric motor in real time through an electronic control system, thereby reducing the driver's operating burden, especially in the low-speed and parking states, providing a more lightweight and comfortable operating experience; currently, the electric power steering system mainly relies on vehicle speed and steering angle to adjust the assist output of the motor; specifically, when driving at low speeds, the system provides a large amount of assistance to reduce the difficulty of the steering operation; while when driving at high speeds, the system reduces the assistance to improve handling stability; however, in the control process of the existing electric power steering system, there is a lack of fine adjustment for changes in actual demand, which may lead to inaccurate assistance effects, affecting the responsiveness of the system and the comfort of operation.

[0003] A torque PID controller is a feedback controller commonly used in automatic control systems, widely used in electric power steering systems to accurately adjust the steering assist torque; its core idea is to dynamically adjust the control signal through three control parameters, namely proportional (P), integral (I), and derivative (D), according to the error between the target torque and the actual torque, so as to achieve precise control of the torque; the proportional term (P) generates a control signal according to the magnitude of the current error, quickly responding to the change in the error; the integral term (I) eliminates the long-term steady-state error by accumulating the historical errors; the derivative term (D) predicts the change trend of the future error, effectively suppressing the overshoot and oscillation of the system; the torque PID controller can adjust the output according to the real-time error change to ensure that the system stably and accurately meets the expected steering torque requirement; however, when facing a complex dynamic system, the PID controller may have problems such as untimely adjustment or improper parameter selection, resulting in overreaction or underreaction of the system.

[0004] The Language Education Optimization Algorithm (LEOA) is a heuristic algorithm based on the language education process. The algorithm consists of three main stages: students select teachers, students learn from each other, and individual practice. In the mathematical model, these stages are respectively simulated as different operations in the learning process. The algorithm solves optimization problems through the simulation of these education processes. Compared with other meta-heuristic algorithms, the performance of LEOA has been verified in the practical application of engineering design. However, when facing complex problems, the algorithm may have a slow convergence speed or fall into a local optimum situation. Summary of the Invention

[0005] In view of the problems existing in the above background technology, the present invention provides a control method for an automatic power steering system, aiming to achieve precise control of vehicle automatic power steering by combining the traditional PID control algorithm and the intelligent optimization algorithm. By real-time collecting vehicle steering angle data and dynamic parameters, calculating the target power steering torque, and using a torque PID controller to generate corresponding control signals according to the error. To further improve the control accuracy and response speed, this application optimizes the control parameters of the torque PID controller through an improved language education algorithm. Among them, the improved algorithm includes using a path backtracking feedback mechanism to improve the position update in the population exploration stage. At the same time, through a dynamic target guidance and gravitational effect optimization strategy, the position update in the population exploitation stage is improved. Finally, by precisely adjusting the output torque of the power steering motor, it is ensured to highly match the vehicle steering demand. This method can significantly improve the response speed and control accuracy of the automatic power steering system.

[0006] To achieve the above object, the present invention adopts a control method for an automatic power steering system, and the specific steps are as follows.

[0007] S1. Real-time collect the current steering angle data of the vehicle through a steering angle sensor.

[0008] S2. Calculate the target power steering torque based on the steering angle data and the current dynamic parameters of the vehicle.

[0009] S3. The torque PID controller generates corresponding control signals according to the error between the target power steering torque and the actual power steering torque.

[0010] S4. The control parameters of the torque PID controller are optimized through an improved language education algorithm. The specific implementation of the improved language education algorithm is: D1. Use a path backtracking feedback mechanism by introducing the cumulative difference of historical paths and an adaptive feedback factor , improve the position update mathematical model in the exploration stage of the algorithm; D2. Use a dynamic target guidance and gravitational effect optimization strategy to improve the position update mathematical model in the development stage of the algorithm by introducing hierarchical target positions and gravitational effect factors , improve the position update mathematical model in the development stage of the algorithm.

[0011] S5. The control signal precisely adjusts the output torque of the assist motor to ensure a high degree of matching with the vehicle steering demand, thereby achieving precise control of the vehicle's automatic power steering.

[0012] Preferably, the mathematical model of the target assist steering torque in S2 is designed by comprehensively considering the coupling effect of the current steering angle and driving speed on the steering demand, and is constructed in a linear superposition manner in terms of structure; the vehicle steering angle adjustment term reflects the driver's input steering intention, and at the same time, a dynamic adjustment term proportional to the vehicle speed is introduced to meet the special requirements for steering stability and response flexibility during high-speed driving; the two terms are respectively adjusted by the torque coefficient and the speed-related coefficient to achieve precise matching of the target assist output.

[0013] Preferably, the torque PID controller control model in S3 uses a typical PID adjustment structure to perform closed-loop adjustment on the difference between the target assist steering torque and the actual output steering torque. The core design concept is to continuously sense the deviation between the target torque and the execution result during dynamic operation, and use the three adjustment links of proportional, integral, and differential to generate control signals synergistically; the proportional link provides a direct response to the current torque error, the integral link accumulates historical deviations to eliminate the torque steady-state offset, and the differential link performs feedforward adjustment based on the torque error change trend to suppress rapid disturbances and system overshoot.

[0014] Preferably, the specific method of introducing the historical path cumulative difference and the adaptive feedback factor in S4 is as follows: D11. The historical path cumulative difference establishes a clear measurement system for the total deviation of the historical trajectory by calculating the absolute difference between the position of each iteration and the current optimal solution of the population, and accumulating it linearly. The specific mathematical model is: ; In the formula, t is the current iteration number, T is the maximum iteration number, is the position of the i-th population individual in the j-th dimension at the t-th iteration, is the current optimal solution in the j-th dimension at the t-th iteration; D12. Adaptive feedback factor By calculating the cumulative difference between the position at the t-th iteration and the current position to represent the trajectory fluctuation amplitude, and then performing scale normalization with the product of the number of iterations and the problem dimension, a normalized mean measure of trajectory fluctuation is constructed, and then embedded into a non-linear inverse proportional mapping function structure. The specific mathematical model is as follows: ; In the formula, is the current position of the i-th population individual in the j-th dimension, m is the problem dimension, and the meanings of other parameters are the same as above.

[0015] Preferably, the mathematical model for position update in the exploration stage of the improved algorithm is: ; In the formula, is the new position of the i-th population individual in the j-th dimension in the exploration stage of the algorithm, is the current position of the i-th population individual in the j-th dimension, r is a random value between 0 and 1, is the position of the selected teacher, I is a value randomly selected from the set {1, 2}, is the adaptive feedback factor, is the cumulative difference of the historical path.

[0016] Preferably, the cumulative difference of the historical path is an index that evaluates the behavior of population individuals by considering the difference between the path traveled by population individuals during the optimization process and the optimal solution; specifically, calculates the deviation degree between population individuals and the optimal solution during multiple iterations. It guides the exploration direction of population individuals by accumulating the difference between the current position of population individuals and the historical optimal solution; by introducing this term, population individuals can adjust their current search behavior according to their positions in historical iterations, avoiding making decisions only based on local information; in this way, when population individuals stagnate in a certain local area, the cumulative difference of the historical path enables population individuals to automatically make position adjustments, thereby breaking through the local optimum and expanding the search range.

[0017] Preferably, the adaptive feedback factor further optimizes the search strategy of population individuals; specifically, by evaluating the cumulative difference of the path, the adaptive feedback factor can intelligently adjust the size of the current step; when population individuals have experienced a large path difference in previous iterations, It will increase, which means that the individuals in the population need to make a greater adjustment to avoid being continuously trapped in the local optimal region, thereby strengthening the global exploration; while when the path differences of the individuals in the population are small, it will decrease. At this time, the individuals in the population tend to refine the search and improve the accuracy of local optimization; through this adaptive adjustment mechanism, the algorithm can flexibly adjust the search strategy according to the historical performance of the individuals in the population, enabling the individuals in the population to effectively conduct global exploration and make fine adjustments when approaching the optimal solution.

[0018] Preferably, in S4, a hierarchical target position and a gravitational effect factor are introduced in the following specific ways: D21. Hierarchical target position Taking the global optimal solution as the basic reference point, using the total relative change in the positions of an individual over the past several consecutive generations, a composite time-domain position reference structure is constructed through non-linear attenuation on the iterative scale. The specific mathematical model is: ; In the formula, is the hierarchical target position of the i-th individual in the population in the j-th dimension at the t-th iteration, is the global optimal solution in the j-th dimension at the t-th iteration, is the current target position offset of the i-th individual in the population in the j-th dimension at the t-th iteration, and t is the current iteration number; D22. Gravitational effect factor Based on the absolute difference between the objective function value of the current individual position and the objective function value of the current optimal solution, after adding 1 to it, an inverse proportion mapping is implemented to form a non-linear compression mapping function of fitness difference. The specific mathematical model is: ; In the formula, is the objective function value of the current individual position in the population, is the objective function value of the current optimal solution.

[0019] Preferably, the mathematical model for position update in the exploitation stage of the improved algorithm is: ; In the formula, is the new position of the i-th individual in the population in the j-th dimension in the exploitation stage of the algorithm, is the current position of the i-th individual in the j-th dimension, is the step size factor, is the upper limit of the algorithm search space, is the lower limit of the algorithm search space, r is a random value between 0 and 1, t is the current iteration number, is the hierarchical target position, is the gravitational effect factor.

[0020] Preferably, the hierarchical target position is introduced to optimize the update of the target position; in the position update mathematical model in the original algorithm development stage, the position update is mainly adjusted based on the difference between the current position of the population individuals and the boundary; in the improved position update mathematical model, by introducing the global optimal solution and the offset of the current target position of the population individuals , the population individuals not only rely on local information during the search process, but also consider the position of the global optimal solution, and guide the population individuals to approach the global optimal solution by dynamically adjusting the target position.

[0021] Preferably, the gravitational effect factor is introduced to dynamically adjust the pace of the population individuals; in the position update mathematical model in the original algorithm development stage, the position update of the population individuals only depends on the fixed step size and boundary value, while in the improved position update mathematical model, the gravitational effect factor is used to adjust the step size according to the difference between the position of the population individuals and the objective function value of the current optimal solution, so that the population individuals can reduce the step size when approaching the global optimal solution, thereby improving the convergence speed.

[0022] Preferably, the control parameters of the torque PID controller in S4 are optimized by an improved language education algorithm, and the specific steps are as follows: step1. Initialize the parameters of the improved language education algorithm, including the population size N, the maximum number of iterations T, the problem dimension m, the upper limit of the search space, and the lower limit of the search space; step2. Encode the control parameters Kp, Ki, and Kd of the torque PID controller as the solutions in the search space of the improved language education algorithm. As the algorithm iterates, the updated position of the population individuals reflects the corresponding control parameters of the torque PID controller; step3. The algorithm iteratively searches for the optimal solution and updates the position of the population individuals; step4. Calculate the fitness value of each population individual, and record the optimal population individual position corresponding to the current minimum fitness value;

[0023] Preferably, in step 2, the control parameters Kp, Ki, and Kd of the torque PID controller are encoded as the solutions of the improved language education algorithm search space. As the algorithm iterates, the positions of the population individuals are updated, that is, the solutions of the algorithm search space are updated, which also means updating the control parameters of the torque PID controller. The encoding vector is as follows: ; In the formula, x is the encoding vector, that is, the solution of the improved language education algorithm search space. Kp, Ki, and Kd are the proportional parameter, integral parameter, and differential parameter of the torque PID controller respectively.

[0024] Preferably, the fitness function model used to calculate the fitness value in step 4 is constructed based on the principle of minimizing the error energy. Taking the deviation between the target assist steering torque and the actual assist steering torque as the core variable, it measures the matching degree between the system output and the reference target during the entire control process. By squaring and accumulating the error over time, the double punishment for the persistence and amplitude of the deviation is strengthened, so that the time periods with large errors occupy a higher weight in the overall evaluation.

[0025] Preferably, the mathematical model of the output torque of the assist motor in S5 is designed by considering three key factors: the output signal of the torque PID controller, the load on the motor, and the system response delay. First, taking the control signal as the basic adjustment amount, and establishing a proportional relationship with the actual output torque through the motor gain coefficient . Subsequently, the load influence coefficient is introduced to dynamically adjust the effective output of the control signal at different load levels, so that when the motor load is close to the rated limit, the output torque will be correspondingly weakened, reflecting the system's inhibitory response to load changes. Finally, the exponential delay term is used to simulate the inertial characteristics of the motor, making the entire output have a more realistic dynamic response process.

[0026] By adopting the above technical solutions, the beneficial effects of the present invention are as follows: Based on the traditional PID control method, the control parameters of the torque PID controller are optimized by introducing an improved language education algorithm, significantly enhancing the intelligence and self-adaptability of the control; specifically, by introducing a path backtracking feedback mechanism and combining the historical path cumulative difference and the adaptive feedback factor, the algorithm can better adjust the position update of the population individuals during the exploration stage, avoiding premature convergence to the local optimal solution, thereby increasing the global search ability; while in the development stage, through the dynamic target guidance and gravitational effect optimization strategy, and combining the hierarchical target position and the gravitational effect factor, the population individuals can dynamically adjust the search pace according to the path difference and the global optimal solution information, further improving the response accuracy and stability of the system; this control method can accurately calculate the required power steering torque according to the real-time steering angle data and vehicle dynamic parameters, and generate a control signal through the torque PID controller to dynamically adjust the output torque of the power assist motor, thereby achieving precise control of the vehicle's automatic power steering. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 It is a flowchart of the steps of a control method for an automatic power steering system.

[0028] Figure 2 It is a flowchart of the specific steps for optimizing the control parameters of the torque PID controller by the improved language education algorithm.

[0029] Figure 3 It is a comparison chart of the fitness value changes during the optimization process of the existing algorithm and the algorithm of the present invention.

[0030] Figure 4 It is a process chart of the optimization process of the Kp parameter value of the torque PID controller by the existing algorithm and the algorithm of the present invention.

[0031] Figure 5 It is a process chart of the optimization process of the Ki parameter value of the torque PID controller by the existing algorithm and the algorithm of the present invention.

[0032] Figure 6 It is a process chart of the optimization process of the Kd parameter value of the torque PID controller by the existing algorithm and the algorithm of the present invention.

[0033] Figure 7 It is a comparison chart of the control effects of the existing method and the method of the present invention on optimizing the torque PID controller of the automatic power steering system. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0034] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts belong to the scope of protection of the present invention.

[0035] The present invention provides a technical solution: a control method for an electric power steering system, specifically including the following steps, as Figure 1 shown.

[0036] S1. Real-time collect the current steering angle data of the vehicle through a steering angle sensor.

[0037] S2. Calculate the target assist steering torque based on the steering angle data and the current dynamic parameters of the vehicle.

[0038] Specifically, the mathematical model for calculating the target assist steering torque in S2 is: ; ; ; ; ; In the formula, is the target assist steering torque, is the torque coefficient, is the current vehicle steering angle, is the vehicle speed related coefficient, is the current vehicle speed, is the original signal output by the steering angle sensor, is the signal processing function, is the maximum output assist torque of the system at the maximum steering angle , is the original vehicle speed signal, is the scale factor, is the vehicle mass, and n is time.

[0039] Specifically, the code for implementing the mathematical model of the target assist steering torque in MATLAB is: M_max = 15; % Maximum output assist torque theta_max = 30; % Maximum steering angle alpha_vehicle = 0.1; % Scale factor m_vehicle = 1500; % Vehicle mass % Read sensor data data = readmatrix('sensor_data.csv'); % Data file, column 1 is the steering angle, column 2 is the vehicle speed theta_vehicle_raw = data(:, 1); % Extract the steering angle data v_raw = data(:, 2); % Extract the vehicle speed data f_sensor_theta = @(y) y; % Process the steering angle f_sensor_speed = @(s) s; % Process the vehicle speed K_steering = M_max / theta_max; % Calculate the torque coefficient M_target = zeros(1, length(theta_vehicle_raw)); for i = 1:length(theta_vehicle_raw) theta = f_sensor_theta(theta_vehicle_raw(i)); % Collect the steering angle v_speed = f_sensor_speed(v_raw(i)); % Collect the vehicle speed M_target(i) = K_steering * theta + (alpha_vehicle * m_vehicle * v_speed); % Calculate the target assist steering torque end

[0040] S3. The torque PID controller generates a corresponding control signal based on the error between the target assist steering torque and the actual assist steering torque.

[0041] Specifically, the mathematical model for the torque PID controller to generate the control signal is as follows: ; ; In the formula, is the target assist steering torque and the error between the actual assist steering torque , Kp, Ki, Kd are the parameters of the torque PID controller, b is the total running time of the system, is the torque error at the a-th moment.

[0042] S4. The control parameters of the torque PID controller are optimized by an improved language education algorithm. The specific implementation of the improved language education algorithm is as follows: D1. Use a path backtracking feedback mechanism to improve the position update mathematical model in the exploration stage of the algorithm by introducing historical path cumulative differences and an adaptive feedback factor . D2. Use a dynamic target guidance and gravitational effect optimization strategy to improve the position update mathematical model in the development stage of the algorithm by introducing hierarchical target positions and a gravitational effect factor .

[0043] Specifically, the mathematical models for introducing historical path cumulative differences and an adaptive feedback factor in S4 are respectively: ; ; where, is the historical path cumulative difference, t is the current iteration number, T is the maximum iteration number, is the position of the i-th population individual in the j-th dimension at the t-th iteration, is the current optimal solution in the j-th dimension at the t-th iteration, is the adaptive feedback factor, is the current position of the i-th population individual in the j-th dimension, and m is the problem dimension; The position update mathematical model in the exploration stage of the improved algorithm is: ; where, is the new position of the i-th population individual in the j-th dimension in the exploration stage of the algorithm, r is a random value between 0 and 1, is the position of the selected teacher, I is a value randomly selected from the set {1, 2}, and the meanings of other parameters are the same as above.

[0044] Specifically, the mathematical models for introducing hierarchical target positions and a gravitational effect factor in S4 are respectively: ; ; where, is the hierarchical target position of the i-th population individual in the j-th dimension at the t-th iteration, is the global optimal solution in the j-th dimension at the t-th iteration, is the current target position offset of the i-th population individual in the j-th dimension at the t-th iteration, is the gravitational effect factor, is the objective function value of the current population individual position, is the objective function value of the current optimal solution, and t is the current iteration number; The position update mathematical model in the exploitation stage of the improved algorithm is: ; In the formula, is the new position of the i-th population individual in the j-th dimension in the exploitation stage of the algorithm, is the current position of the i-th population individual in the j-th dimension, is the step size factor, with a value of 2, is the upper limit of the algorithm search space, is the lower limit of the algorithm search space, r is a random value between 0 and 1, t is the current iteration number, and the meanings of other parameters are the same as above.

[0045] Specifically, the control parameters of the torque PID controller in S4 are optimized by the improved language education algorithm, as Figure 2 shown. The specific steps are as follows: step1. Initialize the parameters of the improved language education algorithm, including the population size N, the maximum iteration number T, the problem dimension m, the upper limit of the search space, and the lower limit of the search space; step2. Encode the control parameters Kp, Ki, and Kd of the torque PID controller as the solutions in the search space of the improved language education algorithm. As the algorithm iterates, the updated population individual positions reflect the corresponding control parameters of the torque PID controller; step3. The algorithm iteratively searches for the optimal solution and updates the population individual positions; step4. Calculate the fitness value of each population individual and record the optimal population individual position corresponding to the current minimum fitness value;

[0046] step5. Judge whether the current iteration number reaches the maximum iteration number. If so, output the population individual position corresponding to the minimum fitness value and parse it as the three optimal control parameters Kp, Ki, and Kd of the torque PID controller. Otherwise, return to execute Step3. ; Wherein, x is the encoding vector, that is, the solution of the improved language education algorithm search space, and Kp, Ki, and Kd are the proportional parameter, integral parameter, and differential parameter of the torque PID controller, respectively.

[0047] Specifically, the fitness function used to calculate the fitness value in step 4 is: ; Wherein, F is the fitness value.

[0048] S5. The control signal precisely adjusts the output torque of the assist motor to ensure a high degree of matching with the vehicle steering demand, thereby achieving precise control of the vehicle's automatic power steering.

[0049] Specifically, the mathematical model of the output torque of the assist motor in S5 is: ; Wherein, is the gain coefficient of the motor, is the load influence coefficient, is the current motor load, is the maximum load of the motor, is the time constant of the motor.

[0050] Specifically, the code for implementing the mathematical model of the output torque of the assist motor in Matlab is: K_motor = 10; % Motor gain coefficient alpha = 0.1; % Load influence coefficient L_max = 100; % Maximum motor load tau = 0.05; % Motor time constant L0 = 10; % Current motor load u = ILEOA_score; % Control signal, ILEOA_score is the output value of the PID control signal obtained after mapping; n_max = length(L); % Maximum time step M_assist = zeros(1, n_max); % Initialize the output torque array for n = 1:n_max M_assist(n) = K_motor * u(n) * (1 - alpha * L0 / L_max) * exp(-tau* n); % Calculate the output torque of the assist motor end.

[0051] Specifically, during the implementation process, the time-domain mathematical model is converted into an s-domain transfer function mathematical model for the Simulink simulation model: ; In the formula, is the transfer function, s is the s-domain time parameter, is set to 10, is set to 0.1, is set to 10, is set to 100, is set to 0.05; the transfer function of the Simulink simulation model is obtained: .

[0052] Furthermore, the mathematical model in the above embodiment is implemented in Matlab and converted into a runnable code form, including: the position update mathematical model for iterative optimization at each stage of the basic language education algorithm and the improved language education algorithm, the target assist steering torque mathematical model, the assist motor output torque mathematical model, the PID controller mathematical model, the fitness function mathematical model, and the main program. The main program is used to implement the mapping between the basic language education algorithm and the improved language education algorithm and the PID controller parameters respectively, and map the fitness function to the objective function of the Simulink simulation model; set the initialization parameters of the algorithm, including the population size N = 30, the maximum number of iterations T = 30, the problem dimension m = 3, the upper limit = [0.5 0.5 0.5] of the algorithm search space, and the lower limit = [0 0 0] of the search space; run the main program to obtain Figures 3 - 7 the simulation result graph.

[0053] Analysis Figure 3 shows that the existing algorithm reaches the optimal fitness value of 0.412 at the 15th iteration, while the algorithm of the present invention reaches the optimal fitness value of 0.069 at the 12th iteration; as the number of iterations increases, the algorithm of the present invention quickly converges to a lower fitness value, while the existing algorithm converges relatively slowly, the process of the fitness value decreasing is relatively gentle, and it tends to be stable after a certain number of generations; the advantages of the algorithm of the present invention compared with the existing algorithm are reflected in the following aspects: First, the algorithm optimization accuracy is higher, with better stability and robustness; Second, the rapid decrease of the fitness value means that the algorithm can find the optimal solution faster, which is crucial for optimizing the control parameters of the torque PID controller, and can ensure that the electric power steering system performs stably and efficiently under various driving conditions.

[0054] According to Figures 4 - 6It can be obtained that the optimal torque PID control parameters obtained by optimizing the algorithm of the present invention are Kp = 0.163, Ki = 0.148, Kd = 0.126, and the optimal torque PID control parameters obtained by optimizing the existing algorithm are Kp = 0.165, Ki = 0.149, Kd = 0.144; inputting the optimal control parameters into the torque PID controller, the optimal control effect of the electric power steering system is obtained; as Figure 7 shown, the running time of the system Simulink simulation model is set to 500 milliseconds. The target input value of the torque PID controller of the electric power steering system at the beginning is set to 3 units. After 300 milliseconds of steady-state operation, the target input value changes to 2 units. It can be seen from the figure that the method of the present invention has better performance compared with the existing method, especially in terms of system response time and stability. The method of the present invention shows significant advantages; although the curve of the existing method converges in the initial stage, there is a long transition period, and the final steady-state error of the system is large, showing obvious overshoot and slow response of the system; while the curve of the method of the present invention shows faster convergence, and has smaller overshoot throughout the process, and finally reaches a faster and stable steady state; this result means that the method of the present invention can significantly improve the dynamic response and control accuracy of the electric power steering system; specifically, the method of the present invention can more effectively optimize the proportional parameter, integral parameter and differential parameter of the torque PID controller, so that the system can quickly adapt and steer stably in the face of complex driving conditions, which is of great significance for improving the steering control performance of the vehicle at high speed and in complex environments.

Claims

1. A control method for an automatic power steering system, characterized in that: The specific steps are: S1, collecting the current steering angle data of the vehicle in real time through the steering angle sensor; S2, calculating the target power steering torque based on the steering angle data and the current dynamic parameters of the vehicle; S3, the torque PID controller generates a corresponding control signal according to the error between the target power steering torque and the actual power steering torque; S4, the control parameters of the torque PID controller are optimized by an improved language education algorithm; the improved language education algorithm is specifically implemented as follows: D1, a path backtracking feedback mechanism is used to introduce the historical path cumulative difference and adaptive feedback factor ,Improved the mathematical model of position update in the exploration phase of the algorithm; D2. Use a dynamic target guidance and gravitational effect optimization strategy to introduce hierarchical target positions and gravitational effect factor ,improved mathematical model of position update in the algorithm development phase; S5. The control signal accurately adjusts the output torque of the power-assisted motor to ensure a high degree of match with the vehicle's steering requirements, thereby achieving precise control of the vehicle's automatic power-assisted steering.

2. The automatic power steering system control method according to claim 1, characterized in that: The mathematical model for calculating the target power steering torque in S2 is: ； In the formula, To target the power steering torque, is the torque coefficient, is the current vehicle steering angle, is the vehicle speed correlation coefficient, is the current speed of the vehicle, and n is the time.

3. The automatic power steering system control method according to claim 2, characterized in that: The S4 introduces the historical path cumulative difference and adaptive feedback factor The specific method is: D11. Historical path cumulative differences By calculating the absolute difference between the position of each iteration and the current optimal solution of the population, a clear historical trajectory deviation measurement system is established in a linear accumulation manner. The specific mathematical model is: ； In the formula, t is the current iteration number, T is the maximum iteration number, is the position of the i-th individual in the j-th dimension in the t-th iteration, is the current optimal solution of the jth dimension in the tth iteration; D12, Adaptive feedback factor The trajectory fluctuation amplitude is represented by calculating the difference between the position of the tth iteration and the current position, and then the scale is normalized by the product of the number of iterations and the problem dimension to form a standardized trajectory fluctuation mean measurement, which is then embedded in the nonlinear inverse proportional mapping function structure. The specific mathematical model is: ； In the formula, is the current position of the i-th individual in the j-th dimension, m is the problem dimension, and the other parameters have the same meanings as above.

4. The automatic power steering system control method according to claim 3, characterized in that: S4 introduces hierarchical target locations and gravitational effect factor The specific method is: D21. Layered target location Taking the global optimal solution as the basic reference point, using the total relative position change of individuals in the past several generations, and implementing nonlinear attenuation on the iteration scale, a composite time domain position reference structure is constructed. The specific mathematical model is: ； In the formula, is the hierarchical target position of the i-th population individual in the j-th dimension at the t-th iteration, is the global optimal solution of the jth dimension in the tth iteration, is the current target position offset of the i-th individual in the j-th dimension in the t-th iteration, where t is the current iteration number; D22, Gravitational Effect Factor Based on the absolute difference between the current individual position objective function value and the current optimal solution objective function value, an inverse proportional mapping is performed after adding the constant 1 to form a nonlinear compression mapping function of fitness difference. The specific mathematical model is: ； In the formula, is the objective function value of the individual position of the current population, is the objective function value of the current optimal solution.

5. The automatic power steering system control method according to claim 4, characterized in that: The control parameters of the torque PID controller in S4 are optimized by an improved language education algorithm, and the specific steps are as follows: Step 1: Initialize the parameters of the improved language education algorithm, including population size N, maximum number of iterations T, problem dimension m, and upper limit of the search space. , the lower limit of the search space ; Step 2: Encode the control parameters Kp, Ki, and Kd of the torque PID controller as the solution of the search space of the improved language education algorithm. As the algorithm iterates, the updated individual positions of the population reflect the control parameters of the corresponding torque PID controller. Step 3: The algorithm iterates and optimizes to update the positions of individuals in the population; Step 4, calculate the fitness value of each population individual, and record the optimal population individual position corresponding to the current minimum fitness value; Step 5. Determine whether the current number of iterations has reached the maximum number of iterations. If so, output the individual position of the population corresponding to the minimum fitness value and analyze it as the three optimal control parameters Kp, Ki, and Kd of the torque PID controller. Otherwise, return to execute Step 3.

6. The automatic power steering system control method according to claim 5, characterized in that: The mathematical model of the output torque of the power assist motor in S5 is: ； In the formula, is the actual output torque of the power assist motor, is the gain coefficient of the motor, is the control signal output by the torque PID controller, is the load influence coefficient, is the current motor load, is the maximum load of the motor, is the time constant of the motor and n is the time.

Images