An antenna fuzzy PID control method and system based on particle swarm optimization

Abstract

This invention discloses a fuzzy PID control method and system for antennas based on particle swarm optimization, comprising: S1: establishing a simulation model of a radio telescope servo system; S2: constructing a parameter encoding vector for a fuzzy PID controller; S3: using a particle swarm optimization algorithm to globally optimize the parameter encoding vector; S4: when a preset termination condition is met, outputting the globally optimal particle position vector and decoding it to obtain the optimal fuzzy PID controller parameter set; S5: deploying the optimal fuzzy PID controller parameter set into the real-time control system of the radio telescope, and adjusting the PID control gain online based on the real-time acquired antenna angle error and its rate of change to drive antenna movement. This invention's radio telescope servo control method based on particle swarm optimization of fuzzy PID parameters solves the problems of existing technologies where controller parameters rely on human experience, are difficult to globally optimize, and have limited adaptive capabilities.

Description

Technical Field

[0001] This invention relates to the field of astronomical instruments and control technology, and in particular to a fuzzy PID control method and system for antennas based on particle swarm optimization. Background Technology

[0002] Classical PID control: a basic control strategy based on error feedback, where the controller output is determined by a proportional (PID) controller. ),integral( ),differential( The system consists of three linearly superimposed components: the proportional element generates an instantaneous response based on the current error magnitude, determining the system's adjustment speed; the integral element eliminates steady-state deviations by accumulating historical errors, ensuring final accuracy; and the derivative element predicts future trends based on the error change rate, providing anticipatory damping to suppress oscillations and improve dynamic stability. This is achieved by tuning three gain parameters (…). PID controllers can achieve stable and accurate control in simple linear systems.

[0003] Digital PID control is a discretized implementation of the classic PID controller in a computer or digital processor. It transforms the continuous-time integral and differential operations into accumulation and difference calculations in the form of difference equations by discretizing and sampling the continuous PID control law. The output control quantity is calculated and updated in each sampling period based on the current discrete error, the sum of historical errors, and the change in error. The core advantages of digital PID are its flexibility and programmability, facilitating the implementation of complex control logic, online parameter adjustment, and communication with a host computer. Furthermore, software filtering can effectively suppress measurement noise and avoid the drift problems inherent in analog circuits.

[0004] Fixed-parameter fuzzy PID control: an intelligent control method that dynamically adjusts the parameters of a classic PID controller through a fuzzy inference mechanism. Its core lies in adjusting the fixed gain (in the classic PID controller) of the fuzzy PID controller. ) was modified to be based on the system's real-time error ( ) and error change rate ( The method involves online adaptive adjustment of variables. First, it predefines the membership functions and fuzzy rule base of the input and output variables based on experience. At runtime, it uses a fuzzy inference engine to adjust the precise values ​​(…). The PID parameters are converted into fuzzy quantities, and then subjected to rule matching and defuzzification operations to calculate the correction amount of the PID parameters in real time. This allows the controller to dynamically change its control characteristics for different operating conditions. For example, it can enhance the proportional action for a faster response when there is a large error, and strengthen the integral action to improve steady-state accuracy when there is a small error.

[0005] Particle Swarm Optimization (PSO) is a global stochastic optimization technique based on swarm intelligence, inspired by the social cooperative foraging behavior of flocks of birds or schools of fish. The algorithm treats each potential solution to the problem as a "particle" in the search space. The swarm of particles flies within the solution space, and each particle dynamically adjusts its speed and direction based on its historical best position and the swarm's global best position. Through iteration, the swarm gradually converges to the optimal solution region. This algorithm is characterized by its intuitive principle, few parameters, fast convergence speed, and ease of implementation. It is particularly adept at handling complex, nonlinear, high-dimensional optimization problems and is widely used in function optimization, neural network training, and controller parameter tuning.

[0006] As a key component of a radio telescope, the antenna servo control system plays a crucial role in precisely controlling the antenna's pointing direction, driving the antenna structure to accurately align with and track distant celestial targets. Its control performance directly determines the quality of the observation data. This system is a typical complex controlled object characterized by nonlinearity, time-varying behavior, large inertia, strong coupling, and various disturbances (such as wind disturbance, friction, and structural vibration).

[0007] Currently, the widely used classical PID control or digital PID control has a simple structure, but it has inherent drawbacks: fixed PID parameters (proportional coefficients) Integral coefficient Differential coefficients It is difficult to maintain optimal performance under all operating conditions of the telescope (such as low-speed stable tracking, high-speed source switching, and wind resistance). Parameter tuning often relies on the engineer's experience, has a long debugging cycle, and cannot be adjusted adaptively online.

[0008] Fuzzy control can simulate human expert experience and handle nonlinearity and uncertainty, but the design of simple fuzzy control rules is complex, and its steady-state accuracy is sometimes inferior to that of PID. Fuzzy PID controllers combine the advantages of both, adjusting PID parameters online through fuzzy inference, but their core—the formulation of membership functions and fuzzy rules—still heavily relies on prior knowledge and trial and error, lacking a systematic global optimization method, and may get stuck in local optima, failing to fully realize the potential of the controller.

[0009] Particle swarm optimization (PSO) is a global optimization algorithm based on swarm intelligence. It boasts advantages such as fast convergence, few parameters, and ease of implementation, making it suitable for solving complex optimization problems. Applying it to parameter optimization in fuzzy PID control holds promise for automatically and intelligently finding optimal or near-optimal fuzzy rules and parameter sets.

[0010] While existing technologies have addressed the insufficient adaptive capability of fixed PID parameters, their core design process relies on manual trial and error, resulting in fundamental limitations such as insufficient optimization, low efficiency, and suboptimal performance. This invention addresses this limitation by introducing a particle swarm optimization algorithm. This algorithm replaces inefficient manual trial and error with a systematic and automated global search, fundamentally improving the design quality and final performance of the fuzzy PID controller. Although existing technologies have researched the use of intelligent algorithms for controller parameter optimization, their systematic application to solving the full-condition adaptive control problem of a specific object—a radio telescope servo system—and fully considering practical engineering constraints (such as motor saturation and speed limitations), remains insufficient. Summary of the Invention

[0011] To address the problems existing in the prior art, the purpose of this invention is to provide an antenna fuzzy PID control method based on particle swarm optimization. This method combines fuzzy PID control with particle swarm optimization algorithms to optimize servo controller parameters in real time, thereby achieving high-precision, adaptive control of the radio telescope servo system. A further objective of this invention is to provide an antenna fuzzy PID control system based on particle swarm optimization that implements the above method.

[0012] To achieve the above objectives, this invention provides an antenna fuzzy PID control method based on particle swarm optimization, comprising the following steps: S1: Establish a simulation model of the radio telescope servo system. The simulation model is used to characterize the dynamic response characteristics of the antenna azimuth and elevation axes under control commands. S2: Construct the parameter encoding vector of the fuzzy PID controller. The parameter encoding vector adopts a hybrid encoding method, which maps the membership function parameters of the fuzzy controller, the fuzzy inference rule base parameters, and the initial gain parameters of the PID controller into a particle position vector of a unified dimension. S3: Use the particle swarm optimization algorithm to perform global optimization on the parameter encoding vector. In each iteration, decode the particle position vector into specific fuzzy PID controller parameters and load them into the simulation model for closed-loop control simulation. Use the tracking error performance index generated by the simulation as the fitness function value to update the individual optimal position and global optimal position of the particle swarm. S4: When the preset termination condition is met, output the global optimal particle position vector and decode to obtain the optimal fuzzy PID controller parameter set; S5: Deploy the optimal fuzzy PID controller parameter set into the real-time control system of the radio telescope, and adjust the PID control gain online based on the real-time acquired antenna angle error and its rate of change to drive the antenna movement.

[0013] Furthermore, the parameter encoding vector in step S2 specifically includes three parts of parameters: The first part is the center position vector of the fuzzy membership function of the input variable; The second part consists of vectorized elements of the fuzzy inference rule matrix, which are used to define PID parameter correction rules under different fuzzy input levels. The third part consists of the output gain scaling factor and the initial values ​​of the PID basic parameters; the hybrid encoding method supports the simultaneous collaborative optimization of the fuzzy logic structure and PID parameter values ​​within a continuous search space.

[0014] Furthermore, the simulation model established in step S1 is a three-closed-loop control model that includes a current loop, a velocity loop, and a position loop; the simulation model integrates the electromagnetic characteristics of the motor, the characteristics of the mechanical transmission mechanism, the nonlinear friction interference, and the wind load disturbance model, which is used to simulate the physical response of the radio telescope under different operating conditions in an offline environment.

[0015] Furthermore, in step S3, the fitness function adopts the time multiplied by the absolute deviation integral criterion; during the simulation process, the error between the angle output by the simulation model and the target trajectory is calculated in real time, and the error is weighted and integrated to guide the particle swarm to converge toward the optimal solution region with the goal of minimizing the fitness function value.

[0016] Furthermore, the online adjustment process in step S5 includes: acquiring the tracking error and error change rate of the antenna in real time; using the membership function and fuzzy rules in the optimal fuzzy PID controller parameter set to perform fuzzy inference on the tracking error and error change rate, and calculating the correction amount of the PID parameters; and updating the current PID control parameters using the correction amount.

[0017] Furthermore, in step S3, the particle swarm optimization algorithm adopts an inertia weight adjustment strategy, maintaining a large inertia weight in the early stage of iteration for global search, and reducing the inertia weight in the later stage of iteration for local precise search.

[0018] A particle swarm optimization-based antenna fuzzy PID control system, the system being used to implement the aforementioned particle swarm optimization-based antenna fuzzy PID control method, the system comprising: The system simulation module is used to construct mathematical models of the electromechanical systems of radio telescopes and provide a virtual environment for the controlled object. The parameter optimization engine module is used to run the particle swarm optimization algorithm, encode the structural parameters and gain parameters of the fuzzy PID controller into particle vectors, and interact with the system simulation module to automatically search for the globally optimal control parameters based on simulation performance. The real-time control module, which has an embedded fuzzy PID control algorithm, is used to receive the globally optimal parameters generated by the parameter optimization engine module and output control commands according to the real-time operating status of the radio telescope. The servo drive execution module, connected to the real-time control module, is used to respond to control commands and drive the antenna's azimuth and elevation axes to rotate.

[0019] Furthermore, the real-time control module is based on a motion control card, which uses multi-axis synchronous control technology to ensure that the control commands for the azimuth and pitch axes are strictly aligned with the position feedback signals in the time domain.

[0020] Furthermore, the servo drive execution module includes an AC servo motor, a reducer, and an absolute encoder; the absolute encoder is used to detect the antenna position in real time and feed it back to the real-time control module to form a closed-loop control.

[0021] Furthermore, it also includes a host computer interaction module, which is used to set the target observation trajectory, display the real-time tracking error curve, and provide an interface for starting and stopping the parameter optimization engine module.

[0022] This invention presents a radio telescope servo control method based on particle swarm optimization of fuzzy PID parameters, which solves the problems of existing technologies where controller parameters rely on human experience, are difficult to optimize globally, and have limited adaptive capabilities. Attached Figure Description

[0023] Figure 1 This is a structural block diagram of the system of the present invention; Figure 2 This is a flowchart illustrating the overall process of the method of the present invention. Figure 3 This is a schematic diagram of the azimuth axis control architecture; Figure 4 Flowchart of the particle swarm optimization algorithm; Figure 5 A schematic diagram of a multi-algorithm simulation model for a servo controller; Figure 6 The diagram shows the control performance analysis of the three controllers at an azimuth of 100°. Figure 7 The diagram shows the control performance analysis of the three controllers during azimuth switching. Detailed Implementation

[0024] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0025] In the description of this invention, it should be noted that the terms "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing the invention and for simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the invention. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0026] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0027] The following combination Figures 1-7 Specific embodiments of the present invention will be described in detail below. It should be understood that the specific embodiments described herein are for illustrative and explanatory purposes only and are not intended to limit the present invention.

[0028] This invention relates to an antenna fuzzy PID control system based on particle swarm optimization, comprising: The system simulation module is used to construct mathematical models of the electromechanical systems of radio telescopes and provide a virtual environment for the controlled object. The parameter optimization engine module is used to run the particle swarm optimization algorithm, encode the structural parameters and gain parameters of the fuzzy PID controller into particle vectors, and interact with the system simulation module to automatically search for the globally optimal control parameters based on simulation performance. The real-time control module, which has an embedded fuzzy PID control algorithm, is used to receive the globally optimal parameters generated by the parameter optimization engine module and output control commands according to the real-time operating status of the radio telescope. The servo drive execution module, connected to the real-time control module, is used to respond to control commands and drive the antenna's azimuth and elevation axes to rotate.

[0029] The system specifically includes: I. Overall System Architecture The overall architecture of the system of this invention consists of two main parts: hardware and software. The hardware components and software architecture work together to form a complete closed-loop control system, enabling the antenna servo control system to achieve high-precision pointing control of the radio telescope antenna.

[0030] 1. Hardware Components To achieve servo control with small overshoot, fast response, and high precision, this system uses the ETH_GAS_N motion control card as the controller. The motion control card integrates the entire servo system into a unified, nanosecond-level synchronized time domain, realizing multi-axis absolute synchronous control. This ensures that the control commands and position feedback for the azimuth and pitch axes are strictly aligned in time, eliminating trajectory errors caused by communication delays and asynchrony in traditional systems. This is crucial for high-speed tracking of moving celestial bodies.

[0031] The hardware structure of the servo control system for a 6-meter aperture telescope is as follows: Figure 1 As shown, the system consists of two coordinated AC servo control systems for azimuth and elevation, mainly including a controller, motor driver, position sensor, limit switches, motor, and reducer. The host computer connects to the motion control card via a network cable, sending position control commands to the driver in the form of pulse counting. The driver drives the motor, and after reduction by the reducer, the azimuth and elevation axes of the telescope rotate. Simultaneously, an absolute encoder detects the azimuth and elevation angles in real time, transmitting the antenna status information to the motion control card via RS485 communication, and finally feeding back to the host computer, forming a complete closed-loop control system to achieve high-precision pointing and tracking control of the telescope.

[0032] 2. Software Architecture To facilitate real-time monitoring of the telescope's status during commissioning and to flexibly control its start and stop, a dedicated host computer software was developed for this system. The software is developed using Python, with a GUI designed using QT Designer, and integrates serial communication functionality for data exchange with the motion control card. The software comprises four main modules: a radio telescope servo system simulation model, a fuzzy PID controller module, a particle swarm optimization algorithm engine module, and a radio telescope control module.

[0033] The radio telescope servo system simulation model is a mathematical model used to accurately simulate the dynamic behavior of a real antenna drive system. The core of this model includes dual azimuth and elevation channels, typically employing a three-closed-loop structure of position, velocity, and current loops. It also integrates mathematical descriptions of the servo motor's electromagnetic characteristics, mechanical transmission mechanism, nonlinear friction, and key external disturbances. This model allows for the reproduction of the telescope's response under different tracking trajectories and interference conditions in a virtual environment. This provides a safe, efficient, and cost-effective testing platform for control algorithm design, parameter tuning, and performance verification, and is an indispensable foundation for offline optimization of fuzzy PID parameters.

[0034] The core function of the fuzzy PID controller module is to base its operation on the real-time tracking error of the antenna. ) and error change rate ( ), dynamically adjust the PID control parameters online ( , , This module first converts the precise input quantities—tracking error e and error change rate ec—into fuzzy linguistic variables through a fuzzification interface. Then, it performs inference based on a pre-defined expert rule base. Finally, it outputs the precise correction amount for the PID parameters through defuzzification. , , This allows the controller to adaptively change its control characteristics and introduce appropriate differential damping according to the error change trend to suppress overshoot, thereby significantly improving the system's adaptability and robustness to nonlinearity, time-varying parameters and external disturbances.

[0035] The particle swarm optimization (PSO) algorithm engine module is the core computational unit for achieving automated global optimization of controller parameters. This module encodes the fuzzy controller parameters to be optimized as "particle" positions in a multi-dimensional space. By simulating the collective intelligent behavior of a flock of birds foraging, it drives the entire particle swarm to collaboratively search within the solution space. In each iteration, each particle dynamically adjusts its flight direction and step size based on its own historical best experience and the globally best experience shared by the group, continuously updating its position. Simultaneously, a servo system simulation model is used to perform closed-loop performance simulation and evaluation of the controller parameters corresponding to each particle position. Guided by the comprehensive fitness function, the entire swarm is guided to efficiently converge to the set of controller parameters with optimal or near-optimal performance, thus completely replacing the traditional parameter tuning process that relies on manual trial and error.

[0036] The radio telescope control module enables local and remote control of the radio telescope, achieving direction and rotation control of the antenna's azimuth and elevation axes. It receives sensor data, collects real-time operating data and tracking errors of the antenna's azimuth and elevation axes, and embeds a fuzzy PID controller with optimal parameters obtained by the particle swarm optimization algorithm engine module. This controller is compared with the target trajectory given by the host computer, and the drive commands are calculated and output to the servo driver in real time to complete the antenna's position servo closed loop.

[0037] II. Core Methods and Steps The core of this invention lies in the systematic integration of the Particle Swarm Optimization (PSO) algorithm with the design process of a fuzzy PID controller, achieving automated and global optimization of servo controller parameters. The following... Figure 2 The key steps in the process shown are explained in detail.

[0038] This invention relates to an antenna fuzzy PID control method based on particle swarm optimization, comprising the following steps: S1: Establish a simulation model of the radio telescope servo system. The simulation model is used to characterize the dynamic response characteristics of the antenna azimuth and elevation axes under control commands. S2: Construct the parameter encoding vector of the fuzzy PID controller. The parameter encoding vector adopts a hybrid encoding method, which maps the membership function parameters of the fuzzy controller, the fuzzy inference rule base parameters, and the initial gain parameters of the PID controller into a particle position vector of a unified dimension. S3: Use the particle swarm optimization algorithm to perform global optimization on the parameter encoding vector. In each iteration, decode the particle position vector into specific fuzzy PID controller parameters and load them into the simulation model for closed-loop control simulation. Use the tracking error performance index generated by the simulation as the fitness function value to update the individual optimal position and global optimal position of the particle swarm. S4: When the preset termination condition is met, output the global optimal particle position vector and decode to obtain the optimal fuzzy PID controller parameter set; S5: Deploy the optimal fuzzy PID controller parameter set into the real-time control system of the radio telescope, and adjust the PID control gain online based on the real-time acquired antenna angle error and its rate of change to drive the antenna movement.

[0039] Specifically: Step 1: Establish a mathematical model of the servo system. Based on the electromechanical parameters of the target radio telescope servo system, establish its dynamic simulation model. This model should include at least two channels: azimuth and pitch. Each channel's model should represent a high-order system including the electromagnetic characteristics of the motor, mechanical transmission mechanism, frictional nonlinearity, and load inertia. (This step is the foundation for all subsequent optimization work; the accuracy of the model directly determines the actual effectiveness of the optimization results.) (1) Construction of control axis control architecture A modular modeling approach was adopted, establishing two independent but structurally identical simulation channels—azimuth and elevation—in the MATLAB / Simulink simulation platform. Each channel was built strictly according to the actual system's three-loop control architecture (current loop, velocity loop, and position loop). Taking the azimuth axis as an example, its three-loop control architecture is as follows: Figure 3 As shown.

[0040] (2) Establishment of mathematical model of servo system The antenna servo system has four mathematical equations, which describe the relevant characteristics of the voltage equation, electromagnetic torque, magnetic flux linkage, and mechanical motion equation. This antenna servo system uses an AC servo motor with rotor magnetic pole orientation, and the motor vector control uses current control with id = 0. Therefore, the mathematical model of the antenna is equivalent to:

[0041] In the formula, , and They are respectively Shaft voltage, current, and inductance, For stator resistance, It is the back electromotive force. It is a permanent magnet flux chain. Let be the angular velocity of the motor shaft rotation. For electromagnetic torque, The number of magnetic pole pairs This represents the total moment of inertia of the motor shaft. The reduction ratio, , , These represent the angular acceleration, angular velocity, and angular displacement of the antenna axis rotation, respectively. This represents the total disturbance along the antenna axis. The torque constant of the motor. For equivalent torsional stiffness, is the damping coefficient.

[0042] Step 2: Design the fuzzy PID controller structure and parameter encoding. Fuzzy PID control adds a fuzzy algorithm to the traditional PID controller. It detects and tracks the error in real time. and error change rate The three parameters of the PID are automatically adjusted using fuzzy rules. , , This allows the system to control the angle more intelligently. The specific process is as follows: (1) The first step is fuzzy logic design: the measured values ​​are converted into fuzzy quantities, the rule base is used to calculate how much the parameters should be adjusted, and finally, the specific values ​​are converted back to control the angle. The following formula gives the fuzzy PID control parameters after tuning and correction: ; In the formula , , and , , They are respectively , , Initial values ​​and scaling factors; , , These are the corresponding parameter increments. The basic universes of discourse for each relevant fuzzy variable in this design are shown in Table 1.

[0043] Table 1. Definition range of fuzzy sets of variables involved in the fuzzy PID controller

[0044] (2) The next step is to formulate fuzzy rules: Based on the expert experience of "rapidly reducing the error when the error is large, and preventing overshoot and improving accuracy when the error is small", an initial fuzzy control rule table is established. The fuzzy rule base is the core of fuzzy control. This design is based on two input quantities (tracking error e, error change rate ec) and three output quantities (PID parameters). , , Based on the correspondence between the rules and the system, and combined with fuzzy mathematics theory and engineering experience, a fuzzy control rule table containing 49 rules was formulated, which fully defines the system's control strategy and behavior criteria, as shown in Table 2.

[0045] Table 2 Three parameters ( , , Fuzzy inference library

[0046] (2) Finally, particle encoding: The particle encoding scheme is the core bridge connecting the design of fuzzy PID controllers and the particle swarm optimization algorithm. Its essence is to transform a controller design problem into a numerical optimization problem. The quality of the encoding directly determines the size of the search space, complexity, and optimization efficiency of the algorithm. This design adopts a fully parameterized, hybrid encoding strategy to ensure the comprehensiveness of the optimization and its engineering feasibility.

[0047] All adjustable parameters of the fuzzy PID controller that have a significant impact on performance are mapped to the position coordinates of a particle in D-dimensional space in the particle swarm optimization algorithm, thus constructing a hybrid parameter particle vector X: ; in, For angle deviation The center vector, Error change rate The center vectors of all vectors have a dimension of 7. It is Matrix, elements Indicates when For the first class, For the first At level, output A single value; It is Matrix, elements Indicates when For the first class, For the first At level, output The single value, corresponding to ; It is Matrix, elements Indicates when For the first class, For the first At level, output The single value, corresponding to All dimensions are 49; , , It is the key gain, which affects the magnitude of the correction, and the dimension is 3; , , The initial values ​​for the PID are all 3 in dimension. Therefore, the particle vector... The total dimension is .

[0048] Step 3: Construct the fitness function. The fitness function is the "compass" of the particle optimization algorithm, and its design must fully reflect the performance requirements of the servo system. Based on the fuzzy control PID algorithm model, the error signal is acquired in real time, and its time absolute deviation integral is used as the fitness function. The calculation formula is as follows: ; in, For error signals, This refers to the time it takes for the control parameters or fitness function to stabilize.

[0049] Step 4: Apply Particle Swarm Optimization (PSO) algorithm for global optimization. This step is the core engine for achieving automated and global design of controller parameters. It runs in an offline simulation environment built on a high-performance computer, iteratively searching for the fuzzy PID controller parameter set that optimizes the overall system performance. This process forms a complete closed loop: "Parameter Proposal (Particles) → Performance Evaluation (Simulation) → Direction Guidance (Fitness) → Parameter Update (PSO)," achieving fully data-driven automated design. The entire process is as follows: Figure 4 As shown.

[0050] (1) Randomly initialize the population, set the solution set range of the gain coefficient, the initial value of the inertia weight, the population size, the velocity of the particles, and the maximum number of iterations, etc.; randomly initialize the position of the particles. ( , , , and ) and speed The initial velocity of each particle exist Randomly generated internally.

[0051] (2) Run the Simulink model based on the fuzzy control PID algorithm, acquire the error signal in real time, and use its time absolute deviation integral as the fitness function. The calculation formula is as follows: ; (3) Calculate the fitness of each particle and update the best fitness of the individual and the particle swarm; if the current particle fitness value is better than its historical best value, then update the individual optimal solution and its corresponding fitness; if it is better than the historical best value of the swarm, then update the global optimal solution and the swarm optimal fitness simultaneously.

[0052] (4) Determine whether the current iteration number meets the termination condition. If it does, stop the iteration and execute step (6); otherwise, execute step (5).

[0053] (5) Update the velocity and position of the particles to obtain a new population; at the same time, return to step (2) to start a new round of iteration. The update formulas for velocity and position are as follows: ; Suppose that the particle swarm has Each particle exist The position in the dimensional search space is represented as Speed ​​is expressed as ,particle The optimal position of an individual is represented as The global optimal position can be represented as .

[0054] in, For particles In the The speed of a dimension in time The value, For particles In the The speed of a dimension in time The value; For particles In the The position of a dimension in time The value, For particles In the The position of a dimension in time The value; and A random number taking values ​​in the range [0,1]. For particles In the The value of the dimension; The optimal position globally is at the 1st The value of the dimension; Inertial weight; and For learning factors.

[0055] (6) Output the global optimal solution, that is, the optimal value of each parameter of the fuzzy control.

[0056] Step 5: Simulation and Analysis of Optimization Results To evaluate the performance of the algorithm of this invention, the designed particle swarm optimization-based fuzzy PID control algorithm was applied to a 6-meter aperture radio telescope. Taking the azimuth axis as an example, a mathematical model was developed to obtain its transfer function. The algorithm controller was constructed using the transfer function, and modeling and simulation of the azimuth axis control of the 6-meter aperture radio telescope were performed. This example uses an azimuth of 100° as the target angle and conducts a simulation experiment lasting 50 seconds. Three control models—traditional PID, fuzzy PID, and particle swarm optimization fuzzy PID algorithm—were established in the MATLAB / Simulink environment. The simulation models are as follows: Figure 5 As shown.

[0057] run Figure 5 The simulation model yielded the following simulation results: Figure 6 As shown in Table 3, compared with traditional PID, PSO-fuzzy PID has a faster response, with a 0.3-second shorter rise time, a 4.5-second shorter settling time, and a 4.59% lower maximum overshoot. Compared with fuzzy PID, its rise time and settling time are reduced by 0.2 seconds and 3.2 seconds respectively, its maximum overshoot is reduced by 1.8%, and its steady-state error is zero. In summary, PSO-fuzzy PID has a faster response speed and better steady-state performance, and can quickly and stably control the azimuth of the radio telescope at 100°. It can be seen that compared with other algorithms, the PSO-fuzzy PID algorithm can better demonstrate its performance, reflecting the adaptability of this algorithm.

[0058] Table 3 Comparison of key indicators for three temperature control schemes at 100° azimuth ; In the daily use of radio telescopes for astronomical observations, angles often need to be switched to point to and track different radio sources. Therefore, an angle switching experiment was conducted to verify the superior performance of the proposed algorithm. For the first 25 seconds of the simulation, the azimuth target angle was set to 100°, and at the 25th second, the target angle was set to 150°. The simulation results of the three algorithms are as follows: Figure 7 As shown.

[0059] Simulation results show that the PSO-fuzzy PID algorithm can reach the target angle more quickly and stably in experiments with different angle switching. This further demonstrates the superiority of this method in updating model parameters.

[0060] Finally, The code is decoded into a specific fuzzy PID controller, and its generalization ability is verified on a simulation model under entirely new test conditions without prior optimization. The optimized membership function parameters and rule matrix are exported. This is then converted into Python code using a code generation tool or through manual programming. The generated controller code is integrated into the telescope's real-time control computer, where the controller executes the following actions during each sampling period: reading encoder feedback and calculating... and Fuzzy reasoning and defuzzification yield results , , The system updates PID parameters, calculates control inputs, and outputs them to the driver. In the initial stages of actual operation, conservative control input limits and safety monitoring logic are set. Based on a small amount of actual observation data, fine-tuning can be performed within the neighborhood of the optimized parameters to achieve the final adaptation from simulation to physical implementation. Specific Implementation

[0061] Antenna status includes four parameters: azimuth angle, elevation angle, rotation speed, and rotation direction. These parameters are obtained by decoding the data returned by the encoder. The specific steps are as follows: Taking a 4096-turn, 1024P azimuth encoder as an example, the communication protocol format for its RS485 interface to send and return data is shown in the table below: Host sends: ; The encoder returned: ; Specifically, when the host computer sends a position query command (01 03 00 09 00 04 94 0B) to the encoder, the encoder returns the following data: ; The table shows that the received data from the azimuth encoder is (01 03 08 00 20 41 7A 00 0107 72 CC). The next step is to decode the received data into the actual azimuth angle of the antenna. The specific decoding steps are as follows: The send instruction indicates that it will read four registers, starting from register 0x0009, which is: Register 0x0009 data is 0x0020; Register 0x000A data is 0x417A; Register 0x000B data is 0x0001; Register 0x000C data is 0x0007; The encoder's total value is 0x0020417A, which is 2113914 in decimal. Therefore, the number of single turns is 2113914 ÷ 1024, and the remainder is 378.

[0062] Then the antenna azimuth angle = .

[0063] The antenna rotation direction value is 0x0001, which represents a clockwise CW direction.

[0064] The antenna rotation speed value is 0x0007, which means the rotation speed is 7 RPM.

[0065] The encoder sampling period for the azimuth axis control system of the radio telescope is... The specific strategy for using PSO-optimized fuzzy PID controller algorithm to achieve stable tracking of radio telescope elevation angle is as follows: Initialization: Set the initial values ​​for the PID parameters, for example... , , .

[0066] Design the initial membership function and rules for the fuzzy controller.

[0067] Initialize the particle swarm: Each particle represents a fuzzy controller.

[0068] Set the sliding window size to 20 sampling points.

[0069] In each control cycle (let's say 0.1 seconds), perform the following steps: a. Read encoder data to obtain the current azimuth angle and calculate the error. =Set value - Current value, Error change rate k represents the sampling point.

[0070] b. Using the current optimal fuzzy controller, input... and ,get , , .

[0071] c. Update PID parameters: , , .

[0072] d. Calculate the control output using the new PID parameters And send it to the actuator orientation driver.

[0073] e. The current and Save to the sliding window.

[0074] f. Calculate the fitness of each particle using the data within the sliding window, and sum the absolute values ​​of the errors within the window.

[0075] g. Run one iteration of PSO, update the velocity and position of each particle, and update the global optimal particle.

[0076] Continue running until the set precision is reached or a stop command is given.

[0077] In this way, the fuzzy controller is continuously optimized using recent historical data, enabling it to adapt to the dynamic changes of the telescope.

[0078] This invention provides a servo control method for radio telescopes based on particle swarm optimization of fuzzy PID parameters. Compared with the prior art, it has achieved significant and multi-level technical progress and positive effects, specifically reflected in the following aspects.

[0079] 1) Achieve automation and global optimization of controller design, eliminating reliance on manual experience.

[0080] Effect: The core design of traditional fuzzy PID controllers—membership functions and fuzzy rule bases—heavily relies on the experience of control engineers and repeated trial and error, resulting in a tedious, subjective process that is prone to getting trapped in local optima. This invention, by introducing particle swarm optimization, completely transforms this design process into an automatically executable numerical optimization problem.

[0081] Specifically, this is manifested in the ability of particle swarm optimization (PSO) to conduct systematic and global exploration within a vast search space consisting of parameters with 167 or even higher dimensions. Through hundreds to thousands of closed-loop simulations, it automatically identifies the optimal parameter combination for overall performance. This completely replaces inefficient and blind manual parameter tuning.

[0082] 2) Significantly improves the overall control performance of the radio telescope servo control system.

[0083] The optimized controller can adjust parameters more precisely, especially in regions where both the error and the rate of change of error are very small. By optimizing the integral action, it effectively suppresses the low-speed "creeping" phenomenon, significantly reduces the root mean square error of the angular position of the radio telescope when tracking the target source, and improves the observation quality of weak radio sources.

[0084] For major external disturbances such as wind, the optimized rule base can generate faster and more appropriate parameter adjustment strategies, making the radio telescope more resistant to angular jitter and tracking deviations caused by wind and enabling it to recover to steady state faster. It also has better adaptability to changes in internal system parameters (such as the moment of inertia that varies with pitch angle and the slow time-varying nature of frictional characteristics), ensuring consistent performance throughout the entire operating range.

[0085] The innovation of this invention does not lie in proposing fuzzy PID control or particle swarm optimization in isolation, but in creatively coupling and systematically integrating the two to construct a complete automated methodology for radio telescope servo systems, from design to optimization. Its core inventive points can be summarized in the following three aspects.

[0086] 1. An automated design paradigm for fuzzy PID controllers, consisting of "full parameter encoding + multi-objective closed-loop evaluation," was proposed. This invention solves the fundamental problem of "how to turn the design of a fuzzy controller into an optimizable problem."

[0087] 1) The system encodes all the key design degrees of freedom of the two-dimensional fuzzy PID controller (including input membership functions, output rule tables, output scaling factors and basic PID parameters) into a high-dimensional real vector. This encoding scheme is the basis for all subsequent automation.

[0088] 2) A multi-objective fitness function was constructed and run on a high-fidelity servo system simulation model. This function comprehensively quantifies the engineering performance of the servo system (ITAE, overshoot, settling time, steady-state error, and control energy consumption). This provides a standard for the "quality" of the controller that can be automatically calculated and objectively measured by a computer.

[0089] 2. A global optimization engine based on an improved particle swarm optimization algorithm was constructed, which enabled efficient exploration of the design space. This invention solves the problem of "how to efficiently and intelligently find the global optimal solution in a huge parameter space".

[0090] The iterative loop of the particle swarm optimization algorithm is deeply integrated with the servo system simulation model. The performance of each particle (i.e., a set of controller parameters) is evaluated by running a complete time-domain simulation, ensuring complete consistency between the optimization objective and the actual dynamic performance.

[0091] 3. A complete technical closed loop of "modeling-coding-optimization-deployment" has been formed to meet the specific needs of servo control for radio telescopes. This invention demonstrates strong pertinence and systematicity.

[0092] 1) All aspects are closely centered around the core challenges of radio telescope servo systems (large inertia, nonlinear friction, strong wind disturbance, and high precision requirements over a wide speed range). The simulation model includes key characteristics such as friction coefficient and wind disturbance torque; the fitness function design reflects the dual requirements of astronomical observations for both low-speed stability and high-speed performance.

[0093] 2) From establishing the system model to finally deploying the optimization results to the real-time controller, a complete, autonomous, and executable solution chain is formed. Users only need to provide basic system parameters to automatically obtain a set of performance-proven, customized optimal fuzzy PID controllers.

[0094] Any process or method described in the flowcharts of this invention or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing a particular logical function or process, which can be implemented in any computer-readable medium for use by an instruction execution system, apparatus, or device. The computer-readable medium can be any medium containing a program for storage, communication, propagation, or transmission for use by an execution system, apparatus, or device, including read-only memory, magnetic disks, or optical disks.

[0095] In the description of this specification, references to terms such as "embodiment," "example," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, those skilled in the art can combine or combine the different embodiments or examples described in this specification and the features therein without causing contradiction.

[0096] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the invention. Therefore, any simple modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention should fall within the protection scope of the present invention.

Claims

1. A fuzzy PID control method for antennas based on particle swarm optimization, characterized in that, Includes the following steps: S1: Establish a simulation model of the radio telescope servo system. The simulation model is used to characterize the dynamic response characteristics of the antenna azimuth and elevation axes under control commands. S2: Construct the parameter encoding vector of the fuzzy PID controller. The parameter encoding vector adopts a hybrid encoding method, which maps the membership function parameters of the fuzzy controller, the fuzzy inference rule base parameters, and the initial gain parameters of the PID controller into a particle position vector of a unified dimension. S3: Use the particle swarm optimization algorithm to perform global optimization on the parameter encoding vector. In each iteration, decode the particle position vector into specific fuzzy PID controller parameters and load them into the simulation model for closed-loop control simulation. Use the tracking error performance index generated by the simulation as the fitness function value to update the individual optimal position and global optimal position of the particle swarm. S4: When the preset termination condition is met, output the global optimal particle position vector and decode to obtain the optimal fuzzy PID controller parameter set; S5: Deploy the optimal fuzzy PID controller parameter set into the real-time control system of the radio telescope, and adjust the PID control gain online based on the real-time acquired antenna angle error and its rate of change to drive the antenna movement.

2. The method according to claim 1, characterized in that, The parameter encoding vector in step S2 specifically includes three parts: The first part is the center position vector of the fuzzy membership function of the input variable; The second part consists of vectorized elements of the fuzzy inference rule matrix, which are used to define PID parameter correction rules under different fuzzy input levels. The third part consists of the output gain scaling factor and the initial values ​​of the PID basic parameters; the hybrid encoding method supports the simultaneous collaborative optimization of the fuzzy logic structure and PID parameter values ​​within a continuous search space.

3. The method according to claim 1, characterized in that, The simulation model established in step S1 is a three-closed-loop control model that includes a current loop, a velocity loop, and a position loop. The simulation model integrates the electromagnetic characteristics of the motor, the characteristics of the mechanical transmission mechanism, the nonlinear friction interference, and the wind load disturbance model, and is used to simulate the physical response of the radio telescope under different operating conditions in an offline environment.

4. The method according to claim 1, characterized in that, In step S3, the fitness function adopts the time multiplied by the absolute deviation integral criterion. During the simulation, the error between the angle output by the simulation model and the target trajectory is calculated in real time, and the error is weighted and integrated to guide the particle swarm to converge toward the optimal solution region with the goal of minimizing the fitness function value.

5. The method according to claim 1, characterized in that, The online adjustment process in step S5 includes: acquiring the tracking error and error change rate of the antenna in real time; using the membership function and fuzzy rules in the optimal fuzzy PID controller parameter set to perform fuzzy inference on the tracking error and error change rate, and calculating the correction amount of the PID parameters; and updating the current PID control parameters using the correction amount.

6. The method according to claim 1, characterized in that, In step S3, the particle swarm optimization algorithm adopts an inertia weight adjustment strategy. In the early stage of the iteration, a large inertia weight is maintained for global search, and in the later stage of the iteration, the inertia weight is reduced for local precise search.

7. A fuzzy PID control system for antennas based on particle swarm optimization, characterized in that, The system is used to implement the method according to any one of claims 1-6, the system comprising: The system simulation module is used to construct mathematical models of the electromechanical systems of radio telescopes and provide a virtual environment for the controlled object. The parameter optimization engine module is used to run the particle swarm optimization algorithm, encode the structural parameters and gain parameters of the fuzzy PID controller into particle vectors, and interact with the system simulation module to automatically search for the globally optimal control parameters based on simulation performance. The real-time control module, which has an embedded fuzzy PID control algorithm, is used to receive the globally optimal parameters generated by the parameter optimization engine module and output control commands according to the real-time operating status of the radio telescope. The servo drive execution module, connected to the real-time control module, is used to respond to control commands and drive the antenna's azimuth and elevation axes to rotate.

8. The system according to claim 7, characterized in that, The real-time control module is based on a motion control card, which uses multi-axis synchronous control technology to ensure that the control commands and position feedback signals of the azimuth and pitch axes are strictly aligned in the time domain.

9. The system according to claim 7, characterized in that, The servo drive execution module includes an AC servo motor, a reducer, and an absolute encoder; the absolute encoder is used to detect the antenna position in real time and feed it back to the real-time control module to form a closed-loop control.

10. The system according to claim 7, characterized in that, It also includes a host computer interaction module, which is used to set the target observation trajectory, display the real-time tracking error curve, and provide an interface for starting and stopping the parameter optimization engine module.

Images