A method for optimizing temperature control in fishery aquaculture ponds

Abstract

The invention discloses a method for optimizing the temperature control of a fishery breeding pond, and belongs to the technical field of PID control optimization. The method comprises the following specific steps: S1, constructing a temperature control system of a fishery breeding pond; S2, improving a dandelion optimization algorithm; S21, using a multi-elite adversarial learning strategy to improve a mathematical model of a descent phase of the dandelion optimization algorithm; S22, using an adaptive segmented search strategy to improve a mathematical model of a landing phase of the dandelion optimization algorithm; S3, optimizing PID controller parameters in the temperature control system of a fishery breeding pond by using the improved dandelion optimization algorithm; S4, using the optimal three parameters Kp, Ki and Kd obtained in S3 for the PID controller of the temperature control system of the fishery breeding pond, and optimizing the control of the temperature of the breeding pond; optimizing the temperature control performance and control precision of the system by optimizing the temperature control system of the fishery breeding pond by using the improved dandelion optimization algorithm.

Description

Technical Field

[0001] The invention belongs to the technical field of PID control optimization, and in particular relates to a method for optimizing temperature control in a fishery aquaculture pond. Background Art

[0002] In aquaculture, water temperature is a key factor affecting the growth, reproduction and health of aquatic organisms. Different species have different suitable water temperature ranges, and water temperature fluctuations will directly affect metabolic rate, immunity and feed utilization. Traditional temperature control methods often rely on manual adjustment or simple constant temperature control, which has problems such as low control accuracy, high energy consumption and slow response, and it is difficult to meet the needs of modern aquaculture for precise environmental control. In response to this challenge, PID control, as a classic feedback control algorithm, has become an ideal choice for temperature control in aquaculture ponds due to its accuracy, fast response and stability. It can not only keep the water temperature within the target range, but also significantly reduce energy consumption and improve aquaculture efficiency.

[0003] PID controller is a classic feedback control algorithm, which consists of three parts: proportional, integral and differential. It is used to adjust the system output to make it close to the target value. Proportional control quickly responds to system deviation through current error, integral control accumulates historical errors to eliminate steady-state deviation, and differential control predicts the error change trend to suppress oscillation and overshoot. The output of PID controller is the weighted sum of these three parts. By adjusting the gain parameters Kp, Ki, Kd of proportional, integral and differential, precise control of the system can be achieved. Due to its simple algorithm, strong adaptability and easy implementation, PID controller is widely used in industrial automation, temperature control, motion control and other fields. However, traditional PID controller has poor adaptability when facing nonlinear, time-varying systems or external interference, and its parameters are difficult to adjust and easily affected by noise, which may lead to decreased control performance or increased energy consumption. Intelligent optimization algorithm can be used to optimize PID control to improve its adaptability and robustness.

[0004] The Dandelion Optimizer (DLO) is an emerging bionic optimization algorithm inspired by the natural process of dandelion seeds spreading and growing in the wind. The algorithm achieves a balance between global search and local optimization by simulating the random diffusion of dandelion seeds under the action of wind, the adaptive screening of seeds after they fall to the ground, and the reproduction process of high-quality populations. Its core mechanisms include seed diffusion (exploration phase) and seed growth (development phase), which can effectively avoid falling into local optimality and are suitable for complex multi-objective optimization problems. The Dandelion Optimizer is widely used in engineering optimization, machine learning, data mining and other fields due to its simplicity, robustness and efficiency. Summary of the invention

[0005] The purpose of the invention is to improve the dandelion optimization algorithm to make it more suitable for the temperature control scenario of fishery breeding ponds. The improved dandelion optimization algorithm has better adaptability and higher precision. By improving the dandelion optimization algorithm, the PID controller parameters in the temperature control system of the fishery breeding pond are optimized. The optimized PID controller can more accurately maintain the water temperature within the target range, reduce the impact of temperature fluctuations on aquatic organisms, and improve the response speed and stability of the system. In addition, the optimization process can also reduce the energy consumption of the control system and avoid frequent start and stop of the equipment, thereby achieving efficient and energy-saving temperature regulation to meet the needs of modern breeding for precise environmental control.

[0006] In order to achieve the above object, the present invention adopts the following technical solution.

[0007] A method for optimizing temperature control in a fishery aquaculture pond, the specific steps are as follows.

[0008] S1. Construct a temperature control system for fishery aquaculture ponds. The system includes: an interactive module, a PID controller, an improved dandelion optimization algorithm module, a temperature adjustment module, and a temperature monitoring module.

[0009] S2. Improve the Dandelion optimization algorithm. The specific improvement strategy is:

[0010] S21. A mathematical model of the descent phase of the dandelion optimization algorithm is improved using a multi-elite adversarial learning strategy. The strategy updates the position of the population by using the top 4 elite individuals in the population as a guide, and explores the solution vector of the symmetric area of ​​the current position by introducing the reverse solution of the current individual.

[0011] S22. An adaptive segmented search strategy is used to improve the mathematical model of the landing phase of the dandelion optimization algorithm. The strategy makes a judgment based on the difference between the optimal individual fitness value and the current individual fitness value, performs segmented updates of the population position, and adaptively adjusts the individual positions of the population in combination with dynamic threshold and step size adjustment.

[0012] S3. Use the improved dandelion optimization algorithm to optimize the PID controller parameters in the temperature control system of the fishery breeding pond, and obtain the optimal set of Kp, Ki, and Kd values ​​through algorithm optimization.

[0013] S4. The optimal three parameters Kp, Ki and Kd obtained in S3 are used in the PID controller of the temperature control system of the fishery breeding pond to optimize the control of the breeding pond temperature.

[0014] Preferably, in said S1, the execution process of each module in the constructed fishery aquaculture pond temperature control system is: first, the target temperature of the aquaculture pond is input through the interactive module, the target temperature and the temperature monitoring module obtain the actual temperature for error calculation, the calculated real-time error e(t) is input into the PID controller, the parameters of the PID controller are optimized by the improved dandelion optimization algorithm module, the optimized PID controller outputs the control quantity u(t), the control quantity u(t) is input into the temperature adjustment module to adjust the temperature of the aquaculture pond, and the specific mathematical model of the temperature adjustment module is:

[0015] (1);

[0016] In formula (1), C1 represents the heat capacity characteristic coefficient of water in the culture pond, C2 represents the heat capacity characteristic coefficient of the culture pond wall and insulation layer, V1 represents the volume of water in the culture pond, V2 represents the volume of the pond wall and insulation layer, T represents the temperature of the culture pond, and K represents the total heat dissipation coefficient. Indicates the ambient temperature, Indicates the heater input power, Represents the cooler input power, where the heater input power and the cooler input power are both controlled by the control quantity u(t). The rated power of the heater or cooler is adjusted by the control quantity u(t), so as to output the specific power required at the corresponding temperature. The calculation formula of u(t) is:

[0017] (2);

[0018] In formula (2), Kp represents proportional gain, Ki represents integral gain, Kd represents differential gain, and e(t) represents real-time error.

[0019] Preferably, in S21, a multi-elite adversarial learning strategy is used to improve the mathematical model of the descending phase of the Dandelion optimization algorithm, and the improved specific mathematical model is:

[0020] (3);

[0021] In formula (3), X(iter+1) represents the updated population position, X(iter) represents the current individual position, ω1 represents the elite weight factor, ω2 represents the opposing position weight factor, rand() represents a random number between [0, 1], ub represents the upper limit of the search space, and lb represents the lower limit of the search space. represents the i-th elite individual in the population, n represents the number of elite individuals selected, and its value is 4.

[0022] Preferably, the multi-elite adversarial learning strategy combines the information of multiple elite individuals, fully utilizes the diversity characteristics of multiple excellent solutions in the population, and guides the population to move towards multiple potential optimal solutions, thereby avoiding the search process from being too concentrated near a single solution, significantly enhancing the efficiency of the global search. At the same time, the adversarial learning strategy effectively avoids the possibility of the algorithm falling into a local optimal solution during the optimization process by generating symmetrical positions of solutions. The combination of the two strategies improves the diversity and quality of population position updates, and improves the convergence speed and optimization accuracy of the algorithm.

[0023] Preferably, in S22, an adaptive segmented search strategy is used to improve the mathematical model of the landing phase of the dandelion optimization algorithm, and the improved specific mathematical model is:

[0024] (4);

[0025] In formula (4), X(iter+1) represents the updated population position, X(iter) represents the current individual position, rand() represents a random number between [0, 1], α and β represent step factors, which are dynamically adjusted according to the ratio between the current number of iterations and the maximum number of iterations. represents the optimal individual position in the population, represents the average value of the first four elite individual positions, and F represents the fitness value of the current position. represents the optimal fitness value, F1 and F2 represent the threshold of fitness difference, Represents the position of a random individual in the population.

[0026] Preferably, the improved landing phase through the adaptive segmented search strategy can segment the population according to the performance of individual fitness values, accurately control the search behavior of different individuals, and set the range of the fitness difference thresholds F1 and F2 to be larger than that in the later stage of the algorithm. In the early stage, the global search ability of the algorithm is improved by a larger value range, and in the later stage, the local search ability of the algorithm is improved by reducing the value range. The step factors α and β are adaptively adjusted according to the number of iterations, which also improves the balance between the global search and the local search of the algorithm. The robustness and stability of the algorithm are enhanced through improvements.

[0027] Preferably, in S3, the improved dandelion optimization algorithm is used to optimize the PID controller parameters in the temperature control system of the fishery aquaculture pond, and the specific steps are:

[0028] S31, initialize the parameters of the improved dandelion optimization algorithm, including the population size 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 initialize the population position through the initial parameters of the algorithm;

[0029] S32. A mapping relationship is established between the improved dandelion optimization algorithm and the PID controller parameters, and the three dimensions of the individual position vector X(iter) of the algorithm correspond to the proportional coefficient Kp, the integral coefficient Ki and the differential coefficient Kd of the PID controller respectively. Specifically, the value X1(iter) of the individual in the first dimension is mapped to Kp, the value X2(iter) of the second dimension is mapped to Ki, and the value X3(iter) of the third dimension is mapped to Kd. Through the iterative update of the optimization algorithm, the position of the individual in the search space is continuously adjusted, so as to optimize the PID parameters to achieve the optimal response of the performance index, and finally obtain the optimal parameter combination of the PID controller;

[0030] S33, setting the fitness value function of the improved dandelion optimization algorithm, calculating the fitness values ​​of the individuals in the initial population through the fitness value function, and sorting them. The specific fitness value function is:

[0031] (5);

[0032] In formula (5), J represents the fitness value, e(t) represents the real-time error value, and L represents the system operation time;

[0033] S34, updating the positions of individuals in the population through the mathematical model of the improved dandelion optimization algorithm, and updating the ranking of individuals in the population according to the fitness values ​​of the updated individuals. The mathematical model of the improved dandelion optimization algorithm is divided into three stages, specifically:

[0034] Step 1, the rising stage, simulates the dandelion seeds rising under the influence of wind, so as to update the position of the population individuals. The specific update formula is shown in formula (6):

[0035] (6);

[0036] In formula (6), X(iter+1) represents the updated population position, X(iter) represents the current individual position, c represents the adjustment factor used to control the step size of seed movement, Vx represents the velocity component of the seed in the x direction, Vy represents the velocity component of the seed in the y direction, Y represents a random variable that obeys the lognormal distribution, rand() represents a random number between [0, 1], Represents a random position in the population, randn represents a random number from a standard normal distribution, iter represents the current iteration number, and max_iter represents the maximum iteration number;

[0037] Step 2, the descending phase, simulates the process of dandelion seeds descending after reaching a certain height. The specific update formula is shown in formula (3);

[0038] Step 3, landing phase. This phase is based on the results of the previous two phases to ensure that the algorithm converges to the global optimal solution. The specific update formula is shown in formula (4);

[0039] S35, determine whether the current number of iterations has reached the maximum number of iterations, if so, complete the optimization and output the optimal solution, otherwise, continue to execute S34 to perform optimization.

[0040] By adopting the above technical scheme, the beneficial effects of the present invention are as follows: the dandelion optimization algorithm is improved by a multi-elite adversarial learning strategy and an adaptive segmented search strategy, thereby improving the overall optimization performance of the algorithm, and increasing the diversity of the population while improving the quality of population position updates. When the algorithm falls into a local optimal solution, it can quickly jump out and continue to optimize. The improved dandelion optimization algorithm is used for parameter optimization of the PID controller of the temperature control system of the fishery breeding pond, thereby improving the stability of the temperature control system, so that the temperature can be stabilized near the target value more quickly, and reducing the impact of temperature fluctuations on aquatic organisms. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 The present invention is a flow chart of a method for optimizing temperature control in fishery aquaculture ponds.

[0042] Figure 2 This is the model diagram of the temperature control system of the fishery breeding pond.

[0043] Figure 3 A comparison chart of the changes in fitness values ​​during the optimization process of the improved dandelion optimization algorithm and the standard dandelion optimization algorithm.

[0044] Figure 4 This is a comparison chart of the effects of optimizing PID controller using the improved dandelion optimization algorithm and the standard dandelion optimization algorithm. DETAILED DESCRIPTION

[0045] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only 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 ordinary technicians in this field without making creative work belong to the scope of protection of the present invention.

[0046] The present invention provides a technical solution: a method for optimizing the temperature control of a fishery aquaculture pond, which specifically comprises the following steps: Figure 1 shown.

[0047] S1. Construct a temperature control system for fishery aquaculture ponds. The system includes: an interactive module, a PID controller, an improved dandelion optimization algorithm module, a temperature adjustment module, and a temperature monitoring module. Figure 2 shown.

[0048] Furthermore, in said S1, the execution process of each module in the constructed fishery aquaculture pond temperature control system is as follows: first, the target temperature of the aquaculture pond is input through the interactive module, the target temperature and the temperature monitoring module obtain the actual temperature for error calculation, the calculated real-time error e(t) is input into the PID controller, the parameters of the PID controller are optimized by the improved dandelion optimization algorithm module, the optimized PID controller outputs the control quantity u(t), the control quantity u(t) is input into the temperature adjustment module to adjust the temperature of the aquaculture pond, and the specific mathematical model of the temperature adjustment module is:

[0049] (1);

[0050] In formula (1), C represents the specific heat capacity of water in the breeding pond, V represents the volume of water in the breeding pond, T represents the temperature of the breeding pond, and K represents the total heat dissipation coefficient. Indicates the ambient temperature, Indicates the heater input power, Represents the cooler input power, where the heater input power and the cooler input power are both controlled by the control quantity u(t), and the calculation formula of u(t) is:

[0051] (2);

[0052] In formula (2), Kp represents proportional gain, Ki represents integral gain, Kd represents differential gain, and e(t) represents real-time error;

[0053] S2. Improve the Dandelion optimization algorithm. The specific improvement strategy is:

[0054] S21. A mathematical model of the descent phase of the dandelion optimization algorithm is improved using a multi-elite adversarial learning strategy. The strategy updates the position of the population by using the top 4 elite individuals in the population as a guide, and explores the solution vector of the symmetric area of ​​the current position by introducing the reverse solution of the current individual.

[0055] S22. An adaptive segmented search strategy is used to improve the mathematical model of the landing phase of the dandelion optimization algorithm. The strategy makes a judgment based on the difference between the optimal individual fitness value and the current individual fitness value, performs segmented updates of the population position, and adaptively adjusts the individual positions of the population in combination with dynamic threshold and step size adjustment.

[0056] Furthermore, in S21, a multi-elite adversarial learning strategy is used to improve the mathematical model of the descending phase of the Dandelion optimization algorithm, and the improved specific mathematical model is:

[0057] (3);

[0058] In formula (3), X(iter+1) represents the updated population position, X(iter) represents the current individual position, ω1 represents the elite weight factor, ω2 represents the opposing position weight factor, rand() represents a random number between [0, 1], ub represents the upper limit of the search space, and lb represents the lower limit of the search space. represents the i-th elite individual in the population, n represents the number of elite individuals selected, and its value is 4.

[0059] Furthermore, in S22, an adaptive segmented search strategy is used to improve the mathematical model of the landing phase of the dandelion optimization algorithm, and the improved specific mathematical model is:

[0060] (4);

[0061] In formula (4), X(iter+1) represents the updated population position, X(iter) represents the current individual position, rand() represents a random number between [0, 1], α and β represent step factors, which are dynamically adjusted according to the ratio between the current number of iterations and the maximum number of iterations. represents the optimal individual position in the population, represents the average value of the first four elite individual positions, and F represents the fitness value of the current position. represents the optimal fitness value, F1 and F2 represent the threshold of fitness difference, Represents the position of a random individual in the population.

[0062] S3. Use the improved dandelion optimization algorithm to optimize the PID controller parameters in the temperature control system of the fishery breeding pond, and obtain the optimal set of Kp, Ki, and Kd values ​​through algorithm optimization.

[0063] Furthermore, in S3, the improved dandelion optimization algorithm is used to optimize the PID controller parameters in the temperature control system of the fishery aquaculture pond, and the specific steps are as follows:

[0064] S31, initialize the parameters of the improved dandelion optimization algorithm, including the population size 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 initialize the population position through the initial parameters of the algorithm;

[0065] S32. A mapping relationship is established between the improved dandelion optimization algorithm and the PID controller parameters, and the three dimensions of the individual position vector X(iter) of the algorithm correspond to the proportional coefficient Kp, the integral coefficient Ki and the differential coefficient Kd of the PID controller respectively. Specifically, the value X1(iter) of the individual in the first dimension is mapped to Kp, the value X2(iter) of the second dimension is mapped to Ki, and the value X3(iter) of the third dimension is mapped to Kd. Through the iterative update of the optimization algorithm, the position of the individual in the search space is continuously adjusted, so as to optimize the PID parameters to achieve the optimal response of the performance index, and finally obtain the optimal parameter combination of the PID controller;

[0066] S33, setting the fitness value function of the improved dandelion optimization algorithm, calculating the fitness values ​​of the individuals in the initial population through the fitness value function, and sorting them. The specific fitness value function is:

[0067] (5);

[0068] In formula (5), J represents the fitness value, e(t) represents the real-time error value, and L represents the system operation time;

[0069] S34, updating the positions of individuals in the population through the mathematical model of the improved dandelion optimization algorithm, and updating the ranking of individuals in the population according to the fitness values ​​of the updated individuals. The mathematical model of the improved dandelion optimization algorithm is divided into three stages, specifically:

[0070] Step 1, the rising stage, simulates the dandelion seeds rising under the influence of wind, so as to update the position of the population individuals. The specific update formula is shown in formula (6):

[0071] (6);

[0072] In formula (6), X(iter+1) represents the updated population position, X(iter) represents the current individual position, c represents the adjustment factor used to control the step size of seed movement, Vx represents the velocity component of the seed in the x direction, Vy represents the velocity component of the seed in the y direction, Y represents a random variable that obeys the lognormal distribution, rand() represents a random number between [0, 1], Represents a random position in the population, randn represents a random number from a standard normal distribution, iter represents the current iteration number, and max_iter represents the maximum iteration number;

[0073] Step 2, the descending phase, simulates the process of dandelion seeds descending after reaching a certain height. The specific update formula is shown in formula (3);

[0074] Step 3, landing phase. This phase is based on the results of the previous two phases to ensure that the algorithm converges to the global optimal solution. The specific update formula is shown in formula (4);

[0075] S35, determine whether the current number of iterations has reached the maximum number of iterations, if so, complete the optimization and output the optimal solution, otherwise, continue to execute S34 to perform optimization.

[0076] S4. The optimal three parameters Kp, Ki and Kd obtained in S3 are used in the PID controller of the temperature control system of the fishery breeding pond to optimize the control of the breeding pond temperature.

[0077] Furthermore, the mathematical model of the dandelion optimization algorithm was improved in Matlab, and the code design and writing of the experiment of the method of the present invention were completed. The maximum number of iterations was initialized to 20, the problem dimension was 3, the population size was 50, and the range of the search space was [100, 0.001]. Simulink was used to construct a simulation model of the temperature control system of the fishery breeding pond, including an interactive module, a PID controller, an improved dandelion optimization algorithm module, a temperature adjustment module, and a temperature monitoring module. The mathematical model of the temperature adjustment module was transformed by Laplace to transform the complex time domain differential equation into a frequency domain algebraic equation that is easier to handle. The converted transfer function was used as the controlled function of the model, and specific parameter values ​​were given. The formula is: , s is a complex variable, run the experimental code and model, the experimental results are as follows Figure 3-Figure 4 shown.

[0078] Furthermore, Figure 3 This is a comparison chart of the fitness value changes during the optimization process of the improved dandelion optimization algorithm and the standard dandelion optimization algorithm. From the figure, it can be seen that the improved dandelion optimization algorithm has a smaller position fitness value during the optimization process, and can reach the vicinity of the optimal solution with a smaller number of iterations, reducing the fluctuations of the system during the adjustment process. Figure 4 This is a comparison chart of the effects of optimizing PID controller using the improved dandelion optimization algorithm and the standard dandelion optimization algorithm. The target value of the PID controller response is set to 1 unit. It can be seen from the figure that the response curve of the PID controller optimized using the improved dandelion optimization algorithm is smoother and the overshoot of the response curve is lower, indicating that the control system has better stability and stronger adaptability.

Claims

1. A method for optimizing temperature control in a fishery aquaculture pond, characterized in that: The specific steps are as follows: S1. Construct a temperature control system for fishery aquaculture ponds. The system includes: an interactive module, a PID controller, an improved dandelion optimization algorithm module, a temperature adjustment module, and a temperature monitoring module; S2. Improve the Dandelion optimization algorithm. The specific improvement strategy is: S21. A mathematical model of the descent phase of the dandelion optimization algorithm is improved using a multi-elite adversarial learning strategy. The strategy updates the position of the population by using the top 4 elite individuals in the population as a guide, and explores the solution vector of the symmetric area of ​​the current position by introducing the reverse solution of the current individual. S22. Using an adaptive segmented search strategy to improve the mathematical model of the landing phase of the dandelion optimization algorithm, the strategy makes a judgment based on the difference between the optimal individual fitness value and the current individual fitness value, performs segmented updates of the population position, and adaptively adjusts the population individual position in combination with dynamic threshold and step size adjustment; S3, using the improved dandelion optimization algorithm to optimize the PID controller parameters in the temperature control system of the fishery aquaculture pond, and obtaining the optimal set of Kp, Ki, and Kd values ​​through algorithm optimization; S4, using the optimal three parameters Kp, Ki, Kd obtained in S3 for the PID controller of the temperature control system of the fishery aquaculture pond to optimize the control of the temperature of the aquaculture pond; Among them, in S21, a multi-elite adversarial learning strategy is used to improve the mathematical model of the descending stage of the dandelion optimization algorithm. The specific mathematical model after improvement is: (3)； In formula (3), X(iter+1) represents the updated population position, X(iter) represents the current individual position, ω1 represents the elite weight factor, ω2 represents the opposing position weight factor, rand() represents a random number between [0, 1], ub represents the upper limit of the search space, and lb represents the lower limit of the search space. represents the i-th elite individual in the population, n represents the number of elite individuals selected, and its value is 4; In S22, an adaptive segmented search strategy is used to improve the mathematical model of the landing phase of the dandelion optimization algorithm. The improved specific mathematical model is: (4)； In formula (4), X(iter+1) represents the updated population position, X(iter) represents the current individual position, rand() represents a random number between [0, 1], α and β represent step factors, which are dynamically adjusted according to the ratio between the current number of iterations and the maximum number of iterations. represents the optimal individual position in the population, represents the average value of the first four elite individual positions, and F represents the fitness value of the current position. represents the optimal fitness value, F1 and F2 represent the threshold of fitness difference, represents the position of a random individual in the population.

2. A method for optimizing temperature control in a fishery aquaculture pond according to claim 1, characterized in that: In S1, the execution process of each module in the temperature control system of the fishery breeding pond is as follows: first, the target temperature of the breeding pond is input through the interactive module, the target temperature and the temperature monitoring module obtain the actual temperature for error calculation, the calculated real-time error e(t) is input into the PID controller, the parameters of the PID controller are optimized by the improved dandelion optimization algorithm module, the optimized PID controller outputs the control quantity u(t), and the control quantity u(t) is input into the temperature adjustment module to adjust the temperature of the breeding pond.

3. A method for optimizing temperature control in a fishery aquaculture pond according to claim 1, characterized in that: In S3, the improved dandelion optimization algorithm is used to optimize the PID controller parameters in the temperature control system of the fishery aquaculture pond, and the specific steps are as follows: S31, initialize the parameters of the improved dandelion optimization algorithm, including the population size 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 initialize the population position through the initial parameters of the algorithm; S32. A mapping relationship is established between the improved dandelion optimization algorithm and the PID controller parameters, and the three dimensions of the individual position vector X(iter) of the algorithm correspond to the proportional coefficient Kp, the integral coefficient Ki and the differential coefficient Kd of the PID controller respectively. Specifically, the value X1(iter) of the individual in the first dimension is mapped to Kp, the value X2(iter) of the second dimension is mapped to Ki, and the value X3(iter) of the third dimension is mapped to Kd. Through the iterative update of the optimization algorithm, the position of the individual in the search space is continuously adjusted, so as to optimize the PID parameters to achieve the optimal response of the performance index, and finally obtain the optimal parameter combination of the PID controller; S33, setting the fitness value function of the improved dandelion optimization algorithm, calculating the fitness values ​​of the individuals in the initial population through the fitness value function, and sorting them. The specific fitness value function is: (5)； In formula (5), J represents the fitness value, e(t) represents the real-time error value, and L represents the system operation time; S34, updating the positions of individuals in the population through the mathematical model of the improved dandelion optimization algorithm, and updating the ranking of individuals in the population according to the fitness values ​​of the updated individuals. The mathematical model of the improved dandelion optimization algorithm is divided into three stages, specifically: Step 1, the rising stage, simulates the dandelion seeds rising under the influence of wind, so as to update the position of the population individuals. The specific update formula is shown in formula (6): (6)； In formula (6), X(iter+1) represents the updated population position, X(iter) represents the current individual position, c represents the adjustment factor used to control the step size of seed movement, Vx represents the velocity component of the seed in the x direction, Vy represents the velocity component of the seed in the y direction, Y represents a random variable that obeys the lognormal distribution, rand() represents a random number between [0, 1], Represents a random position in the population, randn represents a random number from a standard normal distribution, iter represents the current iteration number, and max_iter represents the maximum iteration number; Step 2, the descending phase, simulates the process of dandelion seeds descending after reaching a certain height. The specific update formula is shown in formula (3); Step 3, landing phase. This phase is based on the results of the previous two phases to ensure that the algorithm converges to the global optimal solution. The specific update formula is shown in formula (4); S35, determine whether the current number of iterations has reached the maximum number of iterations, if so, complete the optimization and output the optimal solution, otherwise, continue to execute S34 to perform optimization.

Images