An automated bearing press-fitting force control method

Abstract

The present invention discloses an automatic bearing press-in force control method, belonging to the technical field of PID control optimization. The specific steps are as follows: S1. Construct an automatic bearing press-in force control system; S2. Improve the human memory optimization algorithm. The specific improvement strategy is: S21. Improve the mathematical model in the memory generation stage of the human memory optimization algorithm; S22. Improve the emotional state factor β in the recall behavior stage of the human memory optimization algorithm through an adaptive Gaussian perturbation strategy; S3. Convert the automatic bearing press-in force control system into a mathematical model to be optimized, and optimize the PID controller through the improved human memory optimization algorithm; S4. Apply the optimized parameter values to the automatic bearing press-in force control system to adjust the bearing press-in force and stabilize it near the target pressure value. Through the optimization of the automatic bearing press-in force control system by the improved algorithm, the stability and control performance of the system are enhanced.

Description

Technical Field

[0001] The present invention belongs to the technical field of PID control optimization, and particularly relates to an automatic bearing press-in force control method. Background Art

[0002] The automatic bearing press-in force control technology is an advanced control system integrating sensors, actuators and control algorithms. By real-time monitoring the pressure applied to the bearing and comparing it with the set value, the output of the actuator is automatically adjusted to ensure that the pressure applied during the assembly process always remains within the optimal range, thereby preventing bearing damage or improper installation caused by excessive or too small pressure, improving product quality, reducing failure rates and enhancing production efficiency. It is widely used in the bearing assembly process of industries such as automotive, aviation and machinery manufacturing, and usually uses a PID controller to control the magnitude of the pressure.

[0003] A PID controller is a feedback controller widely used in industrial control systems, aiming to adjust the control input to make the system output reach the set target. It consists of three parts: proportional (P), integral (I) and derivative (D). The proportional part is adjusted according to the current error (the difference between the set value and the actual value), the integral part accumulates past errors to eliminate the steady-state error, and the derivative part predicts future error changes to improve the system response speed. By reasonably adjusting these three parameters, the PID controller can achieve a fast and stable system response and is widely used in various control occasions such as temperature, pressure and flow rate.

[0004] The Human Memory Optimization Algorithm (HMO) simulates human behavior in production, stores preferences for success and failure, mimics the human memory method, and gradually moves towards a better direction and result to find a reasonable optimal solution. Through comparison with other meta-heuristic algorithms on the CEC 2013 test set, the results show that HMO has better optimization ability, and the feasibility of the algorithm is verified through convergence analysis and parameter analysis experiments. Optimizing the parameters of the PID controller with the Human Memory Optimization Algorithm can not only improve the performance of the control system, but also enhance the adaptability and search efficiency of the algorithm, providing an effective solution for complex control problems. Summary of the Invention

[0005] The invention aims to: optimize the PID controller in the automatic bearing press-in force control system using an improved Human Memory Optimization Algorithm, thereby improving the control accuracy and adaptive ability of the control system, quickly finding the optimal parameter settings, reducing the dependence on manual adjustment, enhancing system stability, reducing equipment damage caused by excessive pressure, and enhancing the overall performance, improving the production quality of products, thereby ensuring the correct installation of bearings and improving production efficiency.

[0006] To achieve the above object, the present invention adopts the following technical solutions.

[0007] S1. Construct an automated bearing press - in force control system. The control system is divided into a press - in force controller, a force sensor, a motor drive system, and a data acquisition and pre - processing module. The press - in force controller is composed of an improved human memory optimization algorithm and a PID controller. The PID control is optimized through the improved human memory optimization algorithm to achieve precise control of the pressure. The force sensor real - time collects the press - in force applied during the bearing installation process. The motor drive system serves as the driving force - applying device and adjusts the force application according to the output of the controller. The data acquisition and processing module processes the data of the force sensor and provides feedback to the press - in force controller.

[0008] S2. Improve the human memory optimization algorithm. The specific improvement strategy is as follows:

[0009] S21. Improve the mathematical model in the memory generation stage of the human memory optimization algorithm. By introducing a differential mutation strategy, the population positions updated in the memory generation stage of the original algorithm are differentially mutated, and the mutation intensity is dynamically adjusted through an adaptive mutation probability factor. The specific mutation strategy steps are as follows: First, generate a random judgment factor , if the judgment factor is greater than the set mutation probability factor then perform mutation. If mutation is performed, the fitness value of the individual after mutation needs to be calculated. If the fitness value of the individual after mutation is less than the fitness value of the individual before mutation, then accept the individual after mutation; otherwise, retain the original individual position before mutation.

[0010] S22. Improve the emotional state factor β in the recall behavior stage of the human memory optimization algorithm through an adaptive Gaussian perturbation strategy. It is dynamically adjusted according to the fitness value and the current iteration number during the optimization process, and the emotional state factor is randomly perturbed by adding a Gaussian distribution.

[0011] S3. Convert the automated bearing press - in force control system into an optimization - required mathematical model, and then optimize the PID controller in the automated bearing press - in force control system through the improved human memory optimization algorithm. The optimal set of Kp, Ki, and Kd parameter values are obtained through optimization.

[0012] S4. Apply the optimized parameter values to the automated bearing press - in force control system. The control quantity is output by the optimized PID controller to adjust the bearing press - in force and stabilize it near the target pressure value.

[0013] Preferably, in the step S1, the execution process of the constructed automatic bearing press-in force control system is as follows: First, the set target pressure and the actual pressure obtained by the force sensor are input into the data acquisition and preprocessing module. The pressure value is converted into a corresponding pressure electrical signal, and the error e(t) between the target pressure electrical signal and the actual pressure electrical signal is input into the press-in force controller. The parameters of the PID controller are optimized by improving the human memory optimization algorithm, and the optimal control quantity u(t) is output. The motor drive system adjusts the torque of the motor according to the control quantity and converts the torque into pressure through the rotation system. The mathematical model of the data acquisition and preprocessing module is shown in Equation (1):

[0014] (1);

[0015] In Equation (1), represents the time constant of the system, represents the pressure electrical signal, Ks represents the pressure sensor gain, and F(t) represents the pressure value;

[0016] The motor drive system adjusts the magnitude of the current to adjust the torque Tm of the motor through the control quantity u(t) output by the PID controller optimized by the improved human memory optimization algorithm, and converts the torque of the motor into the pressure value F(t) through the rotation system. The mathematical model of the control quantity u(t) is shown in Equation (2), and the mathematical model of the pressure value F(t) is shown in Equation (3):

[0017] (2);

[0018] In Equation (2), u(t) represents the control quantity, Kp represents the proportional gain, Ki represents the integral gain, Kd represents the differential gain, and e(t) represents the real-time error, , represents the target pressure, represents the actual pressure value;

[0019] (3);

[0020] In Equation (3), F(t) represents the pressure value, Tm represents the motor torque, r represents the equivalent radius of the rotating device, and μ represents the transmission ratio.

[0021] Preferably, in the step S21, for the mathematical model of the memory generation stage of the improved human memory optimization algorithm, the differential mutation strategy is introduced to perform differential mutation on the population position updated in the memory generation stage of the original algorithm, and the intensity of mutation is dynamically adjusted through the adaptive mutation probability factor. The improved mathematical model is as follows:

[0022] (4);

[0023] In Equation (4), X(iter + 1) represents the updated individual position, represents the average position in the population, V represents the function simulating human behavior state, and its calculation formula is V = K • sin(1 / K) • b. K represents a random number between [0.1, 1.3], b represents the iteration factor, b = 2 • (1 - iter / max_iter), iter represents the current iteration number, max_iter represents the maximum iteration number, UB represents the upper limit of the search space, LB represents the lower limit of the search space, rand(1, dim) represents a random number between [1, dim], dim represents the dimension of the optimization problem, represents the mutation probability factor, and its calculation formula is as shown in Equation (5). represents the random judgment factor, X(iter) represents the current individual position, represents the maximum value of the differential mutation intensity, and respectively represent the positions of two randomly selected individuals in the population, represents the average position weighting factor, σ represents the random perturbation factor, and randn(1, daim) represents a random number matrix of a standard normal distribution with dimension dim;

[0024] (5);

[0025] In Equation (5), represents the mutation probability factor, represents the minimum value of the mutation probability, represents the maximum value of the mutation probability, represents the worst fitness value in the population, represents the best fitness value in the population, and f represents the fitness value of the current individual.

[0026] The advantage of this improved strategy is that by introducing differential mutation and a dynamic adjustment mechanism, the mutation probability can be adaptively adjusted according to the fitness value of the current individual, which can effectively balance the relationship between exploration and exploitation in the algorithm optimization process, thereby improving the efficiency and convergence speed of the optimization process. At the same time, by using the average position and random perturbation of the current solution, it can explore more flexibly in the search space, thus avoiding falling into local optimal solutions. Differential mutation also enhances the diversity and adaptability of solutions, making the algorithm perform better in complex problems.

[0027] Preferably, in S22, an adaptive Gaussian perturbation strategy is used to improve the emotional state factor β in the recall behavior stage of the human memory optimization algorithm. The mathematical model of the improved emotional state factor β is:

[0028] (6);

[0029] In formula (6), β represents the emotional state factor, represents the basic value of the emotional state factor, which is used to control the basic intensity of β, represents the average fitness in the population, represents the optimal fitness value in the population, iter represents the current iteration number, max_iter represents the maximum iteration number, and k represents the random adjustment parameter, represents the Gaussian distribution function, represents the mean of the Gaussian distribution, represents the standard deviation of the Gaussian distribution.

[0030] Preferably, in the original algorithm, the β emotional state factor is a random number following a normal distribution. Although it has a certain degree of diversity, it lacks the adaptive ability when dealing with complex problems. After being improved by the strategy in S22, the emotional state factor can adaptively adjust its size according to the current iteration stage and the distribution of the population fitness value. At the same time, combined with the perturbation of the Gaussian distribution, the randomness of the β update is enhanced. This improvement not only improves the update quality of β in the optimization process but also ensures its diversity, thus improving the overall optimization ability of the algorithm.

[0031] Preferably, in S3, the PID controller is optimized by the improved human memory optimization algorithm. The specific steps to obtain the optimal set of Kp, Ki, and Kd parameter values through optimization are as follows:

[0032] S31. Construct a second-order transfer function as the controlled model of the system based on the dynamic characteristics of the automatic bearing pressing force control system. The specific formula is:

[0033] (7);

[0034] In formula (7), G(s) represents the transfer function, Ks represents the gain of the system, s represents the complex frequency domain variable in the Laplace transform, and represent the time constants of the system;

[0035] S32. Initialize the parameters of the improved human memory optimization algorithm, including the population size N, the problem dimension dim, the maximum iteration number max_iter, the upper bound UB of the search space, and the lower bound LB. Generate the initial population positions through the initial parameters of the improved human memory optimization algorithm.

[0036] S33. Map the solution vector [x1, x2, x3] in the improved human memory optimization algorithm to the parameters of the PID controller of the automatic bearing pressing force control system. The dimension dim of the problem to be optimized is 3, that is, x1 corresponds to the proportional gain Kp, x2 corresponds to the integral gain Ki, and x3 corresponds to the derivative gain Kd. In the algorithm optimization, the population position changes continuously, so as to adjust the parameters of the PID control. By continuously adjusting and optimizing the PID, the optimal response is achieved;

[0037] S34. Construct the fitness value function in the improved human memory optimization algorithm. Calculate the fitness values of the individuals in the initial population through the fitness value function, sort them, select the optimal position and the worst position in the population, and add the optimal position and the worst position to the optimal memory matrix and the worst memory matrix respectively. The specific fitness value calculation function is:

[0038] (8);

[0039] In formula (8), J represents the fitness value, T represents the running time of the system, represents the target pressure, represents the actual pressure value;

[0040] S35. Update the positions of the individuals in the population through the mathematical model of the improved human memory optimization algorithm, and update the fitness value sizes of the individuals in the population, the individual fitness value rankings, the worst memory matrix and the optimal memory matrix. The specific steps are as follows:

[0041] step1. First, execute the memory generation stage of the algorithm. The mathematical model of this stage is shown in formula (4). Set the probability value r = 0.3 and generate a random number rand between [0, 1];

[0042] step2. If rand <= r, execute the stage of the human recall failure event. The mathematical model of this stage is shown in formula (9):

[0043] (9);

[0044] In formula (9), X(iter + 1) represents the updated individual position, represents the average position in the population, Xbest represents the optimal position in the population, β represents the emotional state factor, r1, r2 represent random numbers between [0, 1], represents the αth position selected from the worst memory matrix;

[0045] step3. If rand > r and R <= 0.5, execute the stage of the human recall recent event. The mathematical model of this stage is shown in formula (10):

[0046] (10);

[0047] In Equation (10), X(iter + 1) represents the updated individual position, represents the average position in the population, σ and r1 represent random numbers between [0, 1], represents a random individual position in the population, and Pworst represents the worst position at the current iteration;

[0048] Step 4: If rand > 0.5 and rand <= 1 - r, execute the stage where humans search for lost memories. The mathematical model for this stage is shown in Equation (11):

[0049] (11);

[0050] In Equation (11), X(iter + 1) represents the updated individual position, Xbest represents the optimal position in the population, r2 represents a random number between [0, 1], and X(iter) represents the current individual position, represents the optimal position at the current iteration;

[0051] Step 5: Finally, execute the stage of the successful human recollection event. The mathematical model for this stage is shown in Equation (12):

[0052] (12);

[0053] In Equation (12), X(iter + 1) represents the updated individual position, Xbest represents the optimal position in the population, r1 and r2 represent random numbers between [0, 1], β represents the emotional state factor, and its calculation formula is shown in Equation (6), represents the α2 position randomly selected from the optimal memory matrix, represents the average position in the population;

[0054] S36. Determine whether the current iteration number has reached the maximum iteration number. If it has reached the maximum iteration number, complete the optimization and output the optimal solution. Otherwise, continue to execute S35 for optimization.

[0055] By adopting the above technical solutions, the beneficial effects of the present invention are as follows: By improving the human memory optimization algorithm, the adaptability of the algorithm in the optimization process is enhanced, the optimization capabilities in the exploration and exploitation stages of the algorithm are balanced, the convergence speed and optimization accuracy of the algorithm are improved. When the algorithm falls into a local optimal solution, it can quickly jump out and continue to optimize, showing better robustness in dealing with complex problems. The improved human memory optimization algorithm is used to optimize the PID controller in the automatic bearing press-in force control system, thereby improving the control accuracy and adaptive ability of the automatic bearing press-in force control system. This method can quickly find the optimal control parameters, reduce the dependence on manual adjustment, thereby enhancing the stability of the system and reducing the risk of equipment damage caused by excessive pressure. The improvement of the overall control performance helps to improve the production quality of products, ensure the correct installation of bearings, and further improve production efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 FIG. is a flowchart of an automatic bearing press-in force control method.

[0057] Figure 2 FIG. is a model diagram of an automatic bearing press-in force control system.

[0058] Figure 3 FIG. is a flowchart of the execution of the human memory optimization algorithm.

[0059] Figure 4 FIG. is a comparison diagram of the change of fitness values during the optimization process between the improved human memory optimization algorithm and the standard human memory optimization algorithm.

[0060] Figure 5 FIG. is a comparison diagram of the effects of optimizing the PID controller between the improved human memory optimization algorithm and the standard human memory optimization algorithm. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0061] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described 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; based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0062] The present invention provides a technical solution: an automatic bearing press-in force control method, which specifically includes the following steps, as Figure 1 shown.

[0063] S1. Construct an automatic bearing press-in force control system, and the control system is divided into a press-in force controller, a force sensor, a motor drive system, and a data acquisition and preprocessing module, as Figure 2As shown in the figure, the press-in force controller is composed of an improved human memory optimization algorithm and a PID controller. The PID control is optimized by the improved human memory optimization algorithm to achieve precise control of the pressure. The force sensor real-time collects the press-in force applied during the bearing installation process. The motor drive system serves as the driving force application device and adjusts the force application according to the output of the controller. The data acquisition and processing module processes the data of the force sensor and provides feedback to the press-in force controller.

[0064] Further, in the step S1, the execution process of the constructed automated bearing press-in force control system is as follows: First, the set target pressure and the actual pressure obtained by the force sensor are input into the data acquisition and preprocessing module. The pressure value is converted into a corresponding pressure electrical signal, and the error e(t) between the target pressure electrical signal and the actual pressure electrical signal is input into the press-in force controller. The parameters of the PID controller are optimized by the improved human memory optimization algorithm and the optimal control quantity u(t) is output. The motor drive system adjusts the torque of the motor according to the control quantity and converts the torque into pressure through the rotation system. The mathematical model of the data acquisition and preprocessing module is shown in Equation (1):

[0065] (1);

[0066] In Equation (1), represents the time constant of the system, represents the pressure electrical signal, Ks represents the pressure sensor gain, and F(t) represents the pressure value;

[0067] The motor drive system adjusts the magnitude of the current to adjust the torque Tm of the motor through the control quantity u(t) output by the PID controller optimized by the improved human memory optimization algorithm, and converts the torque of the motor into a pressure value F(t) through the rotation system. The mathematical model of the control quantity u(t) is shown in Equation (2), and the mathematical model of the pressure value F(t) is shown in Equation (3):

[0068] (2);

[0069] In Equation (2), u(t) represents the control quantity, Kp represents the proportional gain, Ki represents the integral gain, Kd represents the derivative gain, and e(t) represents the real-time error, , represents the target pressure, represents the actual pressure value;

[0070] (3);

[0071] In Equation (3), F(t) represents the pressure value, Tm represents the motor torque, r represents the equivalent radius of the rotating device, and μ represents the transmission ratio.

[0072] S2. Improve the human memory optimization algorithm. The specific improvement strategy is as follows:

[0073] S21. Improve the mathematical model in the memory generation stage of the human memory optimization algorithm. By introducing the differential mutation strategy, perform differential mutation on the population positions updated in the memory generation stage of the original algorithm, and dynamically adjust the mutation intensity through the adaptive mutation probability factor. The specific mutation strategy steps are as follows: First, generate a random judgment factor , if the judgment factor is greater than the set mutation probability factor then perform mutation. If mutation is performed, the fitness value of the individual after mutation needs to be calculated. If the fitness value of the individual after mutation is less than the fitness value of the individual before mutation, then accept the individual after mutation; otherwise, retain the original individual position before mutation;

[0074] S22. Improve the emotional state factor β in the recall behavior stage of the human memory optimization algorithm through an adaptive Gaussian perturbation strategy, and dynamically adjust it according to the fitness value and the current iteration number in the optimization process. Randomly perturb the emotional state factor by adding a Gaussian distribution.

[0075] Furthermore, in S21, improve the mathematical model in the memory generation stage of the human memory optimization algorithm. By introducing the differential mutation strategy, perform differential mutation on the population positions updated in the memory generation stage of the original algorithm, and dynamically adjust the mutation intensity through the adaptive mutation probability factor. The improved mathematical model is:

[0076] (4);

[0077] In formula (4), X(iter + 1) represents the updated individual position, represents the average position in the population, V represents the function simulating human behavior state, and the calculation formula is V = K • sin(1 / K) • b. K represents a random number between [0.1, 1.3], b represents the iteration factor, b = 2 • (1 - iter / max_iter), iter represents the current iteration number, max_iter represents the maximum iteration number, UB represents the upper limit of the search space, LB represents the lower limit of the search space, rand(1, dim) represents a random number between [1, dim], dim represents the dimension of the optimization problem, represents the mutation probability factor, and the calculation formula is shown in formula (5), represents the random judgment factor, X(iter) represents the current individual position, represents the maximum value of the differential mutation intensity, and respectively represent the positions of two randomly selected individuals in the population, represents the average position weighting factor, σ represents the random perturbation factor, and randn(1, dim) represents a random number matrix of a standard normal distribution with a dimension of dim;

[0078] (5);

[0079] In formula (5), represents the mutation probability factor, represents the minimum value of the mutation probability, represents the maximum value of the mutation probability, represents the worst fitness value in the population, represents the best fitness value in the population, and f represents the fitness value of the current individual.

[0080] Further, in S22, an adaptive Gaussian perturbation strategy is used to improve the emotional state factor β in the recall behavior stage of the human memory optimization algorithm. The mathematical model of the improved emotional state factor β is:

[0081] (6);

[0082] In formula (6), β represents the emotional state factor, represents the basic value of the emotional state factor, which is used to control the basic intensity of β, represents the average fitness in the population, represents the best fitness value in the population, iter represents the current iteration number, max_iter represents the maximum iteration number, and k represents the random adjustment parameter, represents the Gaussian distribution function, represents the mean of the Gaussian distribution, represents the standard deviation of the Gaussian distribution.

[0083] S3. Convert the automated bearing press-in force control system into a mathematical model to be optimized, and then optimize the PID controller in the automated bearing press-in force control system through the improved human memory optimization algorithm to obtain the optimal set of Kp, Ki, and Kd parameter values through optimization.

[0084] Further, in S3, the specific steps to optimize the PID controller through the improved human memory optimization algorithm to obtain the optimal set of Kp, Ki, and Kd parameter values are:

[0085] S31. Construct a second-order transfer function as the controlled model of the system through the dynamic characteristics of the automated bearing press-in force control system. The specific formula is:

[0086] (7);

[0087] In Equation (7), G(s) represents the transfer function, Ks represents the gain of the system, s represents the complex frequency domain variable in the Laplace transform, and represents the time constant of the system;

[0088] S32. Initialize the parameters of the improved human memory optimization algorithm, including the population size N, the problem dimension dim, the maximum number of iterations max_iter, the upper bound UB and the lower bound LB of the search space, and generate the initial population positions through the initial parameters of the improved human memory optimization algorithm;

[0089] S33. Map the solution vector [x1, x2, x3] in the improved human memory optimization algorithm to the parameters of the PID controller of the automatic bearing pressing force control system. The problem dimension dim to be optimized is 3, that is, x1 corresponds to the proportional gain Kp, x2 corresponds to the integral gain Ki, and x3 corresponds to the derivative gain Kd. During the algorithm optimization, the population positions change continuously, thereby adjusting the parameters of the PID control. By continuously adjusting and optimizing the PID, the optimal response is achieved;

[0090] S34. Construct the fitness value function in the improved human memory optimization algorithm, calculate the fitness values of the individuals in the initial population through the fitness value function, sort them, select the optimal position and the worst position in the population, and add the optimal position and the worst position to the optimal memory matrix and the worst memory matrix respectively. The specific fitness value calculation function is:

[0091] (8);

[0092] In Equation (8), J represents the fitness value, T represents the running time of the system, represents the target pressure, represents the actual pressure value;

[0093] S35. Update the positions of the individuals in the population through the mathematical model of the improved human memory optimization algorithm, and update the fitness value sizes of the individuals in the population, the individual fitness value rankings, the worst memory matrix and the optimal memory matrix. The specific steps are as Figure 3 shown:

[0094] step1. First, execute the memory generation stage of the algorithm. The mathematical model of this stage is shown in Equation (4). Set the probability value r = 0.3 and generate a random number rand between [0, 1];

[0095] step2. If rand <= r, execute the stage of the human recall failure event. The mathematical model of this stage is shown in Equation (9):

[0096] (9);

[0097] In Equation (9), X(iter + 1) represents the updated individual position, represents the average position in the population, Xbest represents the optimal position in the population, β represents the emotional state factor, r1 and r2 represent random numbers between [0, 1], represents the α-th position selected from the worst memory matrix;

[0098] Step 3: If rand > r and R <= 0.5, execute the stage where humans recall recent events. The mathematical model of this stage is shown in Equation (10):

[0099] (10);

[0100] In Equation (10), X(iter + 1) represents the updated individual position, represents the average position in the population, σ and r1 represent random numbers between [0, 1], represents a random individual position in the population, Pworst represents the worst position at the current iteration;

[0101] Step 4: If rand > 0.5 and rand <= 1 - r, execute the stage where humans search for lost memories. The mathematical model of this stage is shown in Equation (11):

[0102] (11);

[0103] In Equation (11), X(iter + 1) represents the updated individual position, Xbest represents the optimal position in the population, r2 represents a random number between [0, 1], X(iter) represents the current individual position, represents the optimal position at the current iteration;

[0104] Step 5: If rand > 1 - r, execute the stage where humans recall successful events. The mathematical model of this stage is shown in Equation (12):

[0105] (12);

[0106] In Equation (12), X(iter + 1) represents the updated individual position, Xbest represents the optimal position in the population, r1 and r2 represent random numbers between [0, 1], β represents the emotional state factor, and its calculation formula is shown in Equation (6), represents the α2-th position randomly selected from the optimal memory matrix, represents the average position in the population;

[0107] S36. Determine whether the current iteration count has reached the maximum iteration count. If it has reached the maximum iteration count, complete the optimization and output the optimal solution. Otherwise, continue to execute S35 for optimization.

[0108] S4. Apply the optimized parameter values to the automated bearing press - in force control system. Adjust the bearing press - in force by outputting a control quantity through the optimized PID controller and stabilize it near the target pressure value.

[0109] Furthermore, improve the mathematical model of the human memory optimization algorithm through Matlab, and complete the code design and writing of the experimental method of the present invention. Initialize the parameters of the improved human memory optimization algorithm, with the maximum iteration count max_iter = 30, the population size of 200, and the search space range of [80, 0.001]. Build an automated bearing press - in force control system through the simulation software Simulink. The constructed model includes: a press - in force controller, a force sensor, a motor drive system, and a data acquisition and pre - processing module. The press - in force controller is composed of the improved human memory optimization algorithm and a PID controller. Laplace transform the controlled model of the automated bearing press - in force control system as shown in formula (7). Given specific parameter values according to the actual operating state of the system, the gain Ks of the system is taken as 20, and the time constants and are taken as 5 and 10 respectively. The specific controlled model formula is , set the experimental code in Matlab to combine the algorithm with the constructed automated bearing press - in force control system model, run the code and the model program, and the experimental results are as Figure 4 - 5 shown.

[0110] Furthermore, Figure 4 is a comparison graph of the change of fitness values during the optimization process between the improved human memory optimization algorithm and the standard human memory optimization algorithm. It can be seen from the graph that the improved human memory optimization algorithm has better adaptability during the optimization process. When it falls into a local optimal solution, it can quickly jump out and continue to optimize, and the obtained fitness value is smaller, indicating that the optimized algorithm has higher optimization accuracy and better robustness. Figure 5 is a comparison graph of the effects of optimizing the PID controller between the improved human memory optimization algorithm and the standard human memory optimization algorithm. The target value of the PID controller response value is set to 1 unit. It can be seen from the graph that the PID controller optimized by the improved algorithm has a smaller overshoot and better stability. Compared with the PID controller optimized by the original algorithm, the PID control optimized by the improved algorithm can reach the target value and stabilize in a shorter time, indicating that the overall performance of the automated bearing press - in force control system optimized by the improved human memory optimization algorithm is better.

Claims

1. An automated bearing press force control method, characterized in that: The specific steps are as follows: S1. Construct an automated bearing press force control system, the control system is divided into a press force controller, a force sensor, a motor drive system, and a data acquisition preprocessing module, wherein the press force controller is composed of an improved human memory optimization algorithm and a PID controller, and the PID control is optimized by the improved human memory optimization algorithm to achieve precise control of the pressure, the force sensor collects the press force applied during the bearing installation process in real time, the motor drive system serves as a driving force device, and adjusts the force according to the output of the controller, and the data acquisition and processing module processes the force sensor data to provide feedback for the press force controller; S2. Improve the human memory optimization algorithm. The specific improvement strategies are: S21. Improve the mathematical model of the memory generation phase of the human memory optimization algorithm. By introducing a differential mutation strategy, the population position updated in the memory generation phase of the original algorithm is differentially mutated, and the intensity of the mutation is dynamically adjusted through an adaptive mutation probability factor. The specific mutation strategy steps are as follows: First, generate a random judgment factor , if the judgment factor is greater than the set mutation probability factor Then mutation is executed. If mutation is executed, the fitness value of the individual after mutation needs to be calculated. If the fitness value of the individual after mutation is less than the fitness value of the individual before mutation, the individual after mutation is accepted, otherwise the original individual position before mutation is retained; S22. Improve the emotional state factor β in the recall behavior stage of the human memory optimization algorithm through an adaptive Gaussian perturbation strategy, dynamically adjust it according to the fitness value and the current number of iterations in the optimization process, and randomly perturb the emotional state factor by adding Gaussian distribution; S3, converting the automated bearing pressure-in force control system into a mathematical model to be optimized, and then optimizing the PID controller in the automated bearing pressure-in force control system through an improved human memory optimization algorithm, and obtaining an optimal set of Kp, Ki, and Kd parameter values ​​through optimization; S4. The optimized parameter values ​​are used in the automated bearing pressure control system. The bearing pressure is adjusted by the optimized PID controller output control quantity and stabilized near the target pressure value.

2. The method for controlling the automatic bearing pressing force according to claim 1, characterized in that: In the S1, the execution process of the constructed automated bearing pressure control system is: first, the set target pressure and the actual pressure obtained by the force sensor are input into the data acquisition preprocessing module, the pressure value is converted into the corresponding pressure electrical signal, and the error e(t) between the target pressure electrical signal and the actual pressure electrical signal is input into the pressure controller, the PID controller parameters are optimized by improving the human memory optimization algorithm and the optimal control quantity u(t) is output, and the motor drive system adjusts the torque of the motor according to the control quantity and converts the torque into pressure through the rotation system.

3. The method for controlling the automatic bearing pressing force according to claim 2, characterized in that: In S21, the mathematical model of the memory generation phase of the human memory optimization algorithm is improved, and the population position updated in the memory generation phase of the original algorithm is differentially mutated by introducing a differential mutation strategy, and the intensity of the mutation is dynamically adjusted by an adaptive mutation probability factor. The improved mathematical model is: (4)； In formula (4), X(iter+1) represents the updated individual position, represents the average position in the population, V represents the simulated human behavior state function, and the calculation formula is V=K•sin(1 / K)•b, K represents a random number between [0.1, 1.3], b represents the iteration factor, b=2•(1-iter / max_iter), iter represents the current number of iterations, max_iter represents the maximum number of iterations, UB represents the upper limit of the search space, LB represents the lower limit of the search space, rand(1,dim) represents a random number between [1, dim], and dim represents the dimension of the optimization problem. represents the mutation probability factor, and the calculation formula is shown in formula (5). represents the random judgment factor, X(iter) represents the current individual position, Indicates the maximum value of differential variation intensity, and They represent the positions of two individuals randomly selected from the population, represents the average position weighting factor, σ represents the random perturbation factor, and randn(1, daim) represents a random number matrix of standard normal distribution with dimension dim; (5)； In formula (5), represents the mutation probability factor, represents the minimum value of the mutation probability, represents the maximum value of the mutation probability, represents the worst fitness value in the population, represents the optimal fitness value in the population, and f represents the fitness value of the current individual.

4. The method for controlling the automatic bearing pressing force according to claim 3, characterized in that: In S22, an adaptive Gaussian perturbation strategy is used to improve the emotional state factor β in the recall behavior stage of the human memory optimization algorithm. The mathematical model of the improved emotional state factor β is: (6)； In formula (6), β represents the emotional state factor, Represents the basic value of the emotional state factor, which is used to control the basic strength of β. represents the mean fitness in the population, It represents the optimal fitness value in the population, iter represents the current number of iterations, max_iter represents the maximum number of iterations, and k represents the random adjustment parameter. represents the Gaussian distribution function, represents the mean of the Gaussian distribution, Represents the standard deviation of the Gaussian distribution.

5. The method for controlling the automatic bearing pressing force according to claim 4, characterized in that: In S3, the PID controller is optimized by using an improved human memory optimization algorithm, and the specific steps of obtaining an optimal set of Kp, Ki, and Kd parameter values ​​by searching for the best value are as follows: S31. A second-order transfer function is constructed as the controlled model of the system through the dynamic characteristics of the automatic bearing pressure control system. The specific formula is: (7)； In formula (7), G(s) represents the transfer function, Ks represents the gain of the system, and s represents the complex frequency domain variable in the Laplace transform. and represents the time constant of the system; S32, initializing the parameters of the improved human memory optimization algorithm, including the number of populations N, the problem dimension dim, the maximum number of iterations max_iter, the upper limit UB and the lower limit LB of the search space, and generating the initial population position by improving the initial parameters of the human memory optimization algorithm; S33, mapping the solution vector [x1, x2, x3] in the improved human memory optimization algorithm to the parameters of the PID controller of the automated bearing pressure force control system, the problem dimension to be optimized dim=3, that is, x1 corresponds to the proportional gain Kp, x2 corresponds to the integral gain Ki, and x3 corresponds to the differential gain Kd. The population position changes continuously in the algorithm optimization, thereby adjusting the parameters of the PID control, and achieving the optimal response by continuously adjusting and optimizing the PID; S34, constructing a fitness value function in an improved human memory optimization algorithm, calculating the fitness values ​​of individuals in the initial population through the fitness value function, sorting, selecting the best position and the worst position in the population, and adding the best position and the worst position to the best memory matrix and the worst memory matrix respectively. The specific fitness value calculation function is: (8)； In formula (8), J represents the fitness value, T represents the system operation time, represents the target pressure, Indicates the actual pressure value; S35, updating the positions of individuals in the population by improving the mathematical model of the human memory optimization algorithm, and updating the fitness values ​​of individuals in the population as well as the ranking of individual fitness values, the worst memory matrix and the best memory matrix; S36, determine whether the current number of iterations has reached the maximum number of iterations, if it has reached the maximum number of iterations, complete the optimization and output the optimal solution, otherwise, continue to execute S35 to perform optimization.

Images