A method for optimizing dissolved oxygen control in an aquatic environment of a fishery

Abstract

The application discloses a kind of fishery water body environment dissolved oxygen control optimization method, belong to the technical field of control optimization, including S1, constructs fishery water body management dissolved oxygen PID control system;S2, introduce improved barber optimization algorithm;S3, utilize improved barber optimization algorithm to the dissolved oxygen PID control parameter setting in fishery water body management dissolved oxygen PID control system, and the best PID control parameter is obtained by optimization;S4, three optimal control parameters obtained by using improved barber optimization algorithm optimization are set as the parameter of dissolved oxygen PID controller in fishery water body management dissolved oxygen PID control system, and the effect of dissolved oxygen regulation control is optimized.The application introduces improved barber optimization algorithm, higher optimization accuracy is realized, and the PID controller parameter solution with better performance is obtained, so that water body dissolved oxygen concentration can be maintained more accurately and stably, and reliable technical support is provided for fishery water body management dissolved oxygen.

Description

Technical Field

[0001] This invention belongs to the technical field of control optimization, and in particular relates to a method for controlling and optimizing dissolved oxygen in fishery aquatic environments. Background Technology

[0002] In modern high-density aquaculture, dissolved oxygen (DO) concentration is a key environmental factor determining the success or failure of the aquaculture. Too low a dissolved oxygen level directly leads to oxygen deprivation and death in aquatic organisms, while too high a level increases energy consumption and may trigger gas bubble disease. The dissolved oxygen system in aquaculture waters is a complex dynamic process, characterized by: long response times for large water bodies to aeration, time required for oxygen to diffuse throughout the water body after aeration equipment is turned on, susceptibility to disturbance, and nonlinearity. PID controllers are widely used in industrial process control due to their simple structure, robustness, and high reliability. The performance of a PID controller largely depends on the tuning of the proportional gain Kp, integral time Ki, and derivative time Kd. However, traditional PID parameter tuning methods, such as the Ziegler-Nichols empirical formula method or trial and error method, have significant limitations. They usually rely on the engineer's experience, and for complex industrial objects, the tuning process is time-consuming and laborious, and it is difficult to guarantee the optimal control effect. In particular, for controlled objects with complex dynamic characteristics such as nonlinearity, large time delay, and non-minimum phase, traditional tuning methods often only yield a set of barely usable or even poor-performing parameters, resulting in slow system response, large overshoot, and continuous oscillation, which cannot meet the high-precision and high-stability control requirements of modern industry. Summary of the Invention

[0003] To overcome the technical problems described in the background section, this invention provides an optimization method for dissolved oxygen control in fishery waters. By introducing an improved barber optimization algorithm, it effectively solves the problems of traditional optimization algorithms easily getting trapped in local optima, insufficient optimization accuracy, and poor control effect when facing controlled objects with complex characteristics such as large time delays and non-minimum phases. It also achieves higher optimization accuracy and obtains PID controller parameter solutions with better performance.

[0004] The technical solution of this invention is: a method for controlling and optimizing dissolved oxygen in fishery waters, comprising the following steps:

[0005] S1. Construct a dissolved oxygen PID control system for fishery water body management, including a dissolved oxygen error calculation module, a dissolved oxygen PID controller module, an improved barber optimization algorithm module, a dissolved oxygen regulation module, and a dissolved oxygen monitoring module;

[0006] S2. Introduce an improved barber optimization algorithm. The specific improvement strategy is as follows:

[0007] S201. In the exploration phase, a hybrid guidance ratio parameter Cr is introduced. If the random number is less than Cr, a hybrid guidance strategy combining the global optimal solution and the random solution is adopted; otherwise, the original strategy is used.

[0008] S202. In the utilization phase, an elite-guided precise pruning strategy is adopted. The new position update formula includes a learning part towards the global optimal solution and a random perturbation part that decays with iteration.

[0009] S203. Introduce a stagnation detection and population reset mechanism; when the global optimal solution is continuous... When not updated, proportionally Reset inferior individuals;

[0010] S3. The dissolved oxygen PID control parameters in the dissolved oxygen PID control system of fishery water body management are tuned using the improved barber optimization algorithm, and the optimal control parameters are obtained through optimization. , , ;

[0011] S4. The three optimal control parameters obtained by using the improved barber optimization algorithm are set as the parameters of the dissolved oxygen PID controller in the dissolved oxygen PID control system for fishery water body management, thereby optimizing the dissolved oxygen regulation and control effect.

[0012] Furthermore, in the dissolved oxygen PID control system for fishery water management constructed in step S1, the actual dissolved oxygen value is collected by the dissolved oxygen monitoring module and transmitted to the dissolved oxygen error calculation module. The dissolved oxygen error calculation module receives the set target dissolved oxygen, calculates the error between the target dissolved oxygen and the actual dissolved oxygen, and outputs a real-time error signal to the dissolved oxygen PID controller module. The improved barber optimization algorithm IBaOA module optimizes the internal parameters of the PID controller module. , , The system continuously optimizes control performance. The optimized PID controller module calculates the control quantity based on the error signal and then outputs it to the dissolved oxygen regulation module to perform the dissolved oxygen concentration regulation operation.

[0013] Furthermore, in the exploration phase of step S201, a guiding solution is generated for each current individual to guide its position update. , guide solution The generation strategy is based on a hybrid guided scaling parameter. The decision, and the specific steps include:

[0014] C2011, A random number drawn from the interval [0,1] is less than... At that time, a hybrid bootstrapping strategy is used to generate boot solutions. The specific formula is as follows:

[0015] ,

[0016] in This is the generated guiding solution. The weights are random weights within the interval [0,1]. This is the current globally optimal solution. An individual randomly selected from the population;

[0017] C2012, A random number drawn from the interval [0,1] is not less than... When using a conventional strategy to generate a guiding solution, a guiding solution is randomly selected from all solutions in the population with fitness values ​​better than the current individual. If no better solution exists, then an individual is randomly selected from the entire population (excluding itself) as the guiding solution. .

[0018] Furthermore, in the utilization phase of step S202, the improved algorithm employs a new position update formula:

[0019] ,

[0020] in To determine the learning step size towards the global optimum, This is a random disturbance term.

[0021] Furthermore, the formula for calculating the learning step size from the global optimal solution is:

[0022] ,

[0023] in Guiding learning factors for elites It is a random vector in the interval [0,1].

[0024] Furthermore, the random disturbance term The calculation formula is:

[0025] ,

[0026] in The learning factor is a random perturbation. This represents the current iteration number. The maximum number of iterations, It is a random vector in the interval [0,1].

[0027] Furthermore, a stagnation detection and population reset mechanism is added in step S203, including the following steps:

[0028] C2031. After each iteration, check the global optimal solution update state. If the optimal fitness value has not improved continuously, then the stall counter is activated. Increase by 1;

[0029] C2032, when When the reset mechanism is triggered, the results are sorted in descending order of fitness values, and the top-ranked results are selected. One inferior individual, among which Population size;

[0030] C2033, according to the reset formula Reinitialize the positions of inferior individuals and evaluate their fitness, where lb is the lower bound of the search space, ub is the upper bound of the search space, dim is the problem dimension and takes a value of 3, and the rand function is used to generate a random vector with 1 row and dim columns and each element is uniformly distributed in the interval (0,1).

[0031] C2034. After resetting, update the global optimal solution and clear the stall counter. .

[0032] Furthermore, in step S3, the specific steps for optimizing the PID controller parameters using the improved barber optimization algorithm are as follows:

[0033] S301. Initialize the core parameters of the improved barber optimization algorithm, including population size. Problem dimensions Maximum number of iterations lower limit of search space and upper limit Hybrid guiding ratio Stagnation tolerance limit Reset ratio And randomly generate the initial population location in three-dimensional space;

[0034] S302. Establish the mapping relationship between the algorithm solution space and the PID controller parameters. The three components of the individual position vector correspond to the proportional coefficients of the PID controller. Integral coefficient Differential coefficients ;

[0035] S303. Define the fitness function for evaluating PID control performance. Its mathematical expression is:

[0036] ,

[0037] in The integral of time multiplied by the absolute error. The overshoot of the system response. This is an overshoot penalty item. This is the upper limit of overshoot. The settling time of the system response. As a time-stability penalty item, , , These are the weighting coefficients;

[0038] S304. Execute the core update mechanism of the algorithm, iterate in a loop, and in each iteration, perform the exploration phase update and the utilization phase update for each individual in sequence, and execute the stagnation detection and population reset mechanism at the end of each iteration.

[0039] S305, When the number of iterations reaches... The optimization process terminates at a certain point, and the globally optimal individual position vector is output as the optimal parameter combination for the PID controller. , , .

[0040] Furthermore, in step S4, the specific process of applying the optimized parameters to the control system is as follows:

[0041] S401. When the preset periodic tuning time is met or a significant drop in system performance is detected, the improved barber optimization algorithm module is automatically triggered to start the parameter optimization process.

[0042] S402, Improved Barber Optimization Algorithm Module Based on Fitness Function Evaluate individual fitness and iteratively update the population through exploration, exploitation, and stagnation reset strategies;

[0043] S403. When the algorithm reaches the termination condition, extract the three component values ​​sequentially from the final output global optimal individual position vector, and parse them into proportional gain parameters respectively. Integral coefficient and differential coefficients The specific value;

[0044] S404. Through the system parameter configuration interface, the parsed data is... , and The numerical value is written to the corresponding parameter register of the dissolved oxygen PID controller module so that the controller can immediately use the new parameters for optimized control.

[0045] The beneficial effects of this invention due to the adoption of the above-mentioned technology are as follows: By setting up a hybrid guided exploration and stagnation reset mechanism, this invention achieves higher optimization accuracy and faster convergence speed, and finds a PID controller parameter solution that is superior to the barber optimization algorithm in terms of comprehensive evaluation indicators. In particular, by identifying a differential coefficient Kd that is much larger than the conventional value, the PID controller is directly endowed with strong damping characteristics and predictive ability. When facing controlled objects with complex characteristics such as large time delay and initial reverse response, the optimized closed-loop system exhibits smaller overshoot and weak oscillation, more stable response, and stronger robustness, thereby being able to maintain the dissolved oxygen concentration in the water more accurately and stably, providing a reliable technical guarantee for dissolved oxygen management in fishery water bodies. Attached Figure Description

[0046] Figure 1 This is a flowchart illustrating the present invention.

[0047] Figure 2 This is a flowchart illustrating the application of the optimized parameters of this invention to the control system.

[0048] Figure 3 This is a model diagram of the dissolved oxygen PID control system for fishery water management according to the present invention.

[0049] Figure 4 This is a comparison chart of the optimal fitness values ​​of the improved barber optimization algorithm and the original barber optimization algorithm.

[0050] Figure 5 This is a comparison chart of the optimization parameters Kp between the improved barber optimization algorithm and the barber optimization algorithm of this invention.

[0051] Figure 6 This is a comparison chart of the Ki parameters for the improved barber optimization algorithm and the barber optimization algorithm of this invention.

[0052] Figure 7 This is a comparison chart of the optimization parameters of the improved barber optimization algorithm and the barber optimization algorithm of this invention.

[0053] Figure 8 This is a comparison chart of the system responses of the improved barber optimization algorithm and the original barber optimization algorithm. Detailed Implementation

[0054] Example 1: As Figure 1 As shown, the present invention provides a method for optimizing dissolved oxygen control in fishery waters, comprising the following steps:

[0055] S1. Construct a dissolved oxygen PID control system for fishery water body management, including a dissolved oxygen error calculation module, a dissolved oxygen PID controller module, an improved barber optimization algorithm module, an aeration pump module, and a dissolved oxygen probe module;

[0056] S2. Introduce an improved barber optimization algorithm. The specific improvement strategy is as follows:

[0057] S201. In the exploration phase, a hybrid guidance ratio parameter Cr is introduced. If the random number is less than Cr, a hybrid guidance strategy combining the global optimal solution and the random solution is adopted; otherwise, the original strategy is used.

[0058] S202. In the utilization phase, an elite-guided precise pruning strategy is adopted. The new position update formula includes a learning part towards the global optimal solution and a random perturbation part that decays with iteration.

[0059] S203. Introduce a stagnation detection and population reset mechanism; when the global optimal solution is continuous... When not updated, proportionally Reset inferior individuals;

[0060] S3. The dissolved oxygen PID control parameters in the dissolved oxygen PID control system of fishery water body management are tuned using the improved barber optimization algorithm, and the optimal control parameters are obtained through optimization. , , ;

[0061] S4. The three optimal control parameters obtained by using the improved barber optimization algorithm are set as the parameters of the dissolved oxygen PID controller in the dissolved oxygen PID control system for fishery water body management, thereby optimizing the dissolved oxygen regulation and control effect.

[0062] In the dissolved oxygen PID control system for fishery water management constructed in step S1, the actual dissolved oxygen value is collected by the dissolved oxygen probe module and transmitted to the dissolved oxygen error calculation module. The dissolved oxygen error calculation module receives the set target dissolved oxygen, calculates the error between the target dissolved oxygen and the actual dissolved oxygen, and outputs a real-time error signal to the dissolved oxygen PID controller module. The improved barber optimization algorithm IBaOA module optimizes the internal parameters of the PID controller module. , , The system continuously optimizes control performance. The optimized PID controller module calculates the control quantity based on the error signal and then outputs it to the aeration pump module to perform the dissolved oxygen concentration adjustment operation.

[0063] In the exploration phase of step S201, a guiding solution is generated for each current individual to guide its position update. , guide solution The generation strategy is based on a hybrid guided scaling parameter. The decision, and the specific steps include:

[0064] C2011, A random number drawn from the interval [0,1] is less than... At that time, a hybrid bootstrapping strategy is used to generate boot solutions. The specific formula is as follows:

[0065] ,

[0066] in This is the generated guiding solution. The weights are random weights within the interval [0,1]. This is the current globally optimal solution. An individual randomly selected from the population;

[0067] C2012, A random number drawn from the interval [0,1] is not less than... When using a conventional strategy to generate a guiding solution, a guiding solution is randomly selected from all solutions in the population with fitness values ​​better than the current individual. If no better solution exists, then an individual is randomly selected from the entire population (excluding itself) as the guiding solution. .

[0068] In the utilization phase of step S202, the improved algorithm adopts a new position update formula:

[0069] ,

[0070] in To determine the learning step size towards the global optimum, This is a random disturbance term.

[0071] The formula for calculating the learning step size from the global optimum is:

[0072] ,

[0073] in Guiding learning factors for elites It is a random vector in the interval [0,1].

[0074] random disturbance term The calculation formula is:

[0075] ,

[0076] in The learning factor is a random perturbation. This represents the current iteration number. The maximum number of iterations, It is a random vector in the interval [0,1].

[0077] Step S203 adds a stagnation detection and population reset mechanism, including the following steps:

[0078] C2031. After each iteration, check the global optimal solution update state. If the optimal fitness value has not improved continuously, then the stall counter is activated. Increase by 1;

[0079] C2032, when When the reset mechanism is triggered, the results are sorted in descending order of fitness values, and the top-ranked results are selected. One inferior individual, among which Population size;

[0080] C2033, according to the reset formula Reinitialize the positions of inferior individuals and evaluate their fitness, where lb is the lower bound of the search space, ub is the upper bound of the search space, dim is the problem dimension and takes a value of 3, and the rand function is used to generate a random vector with 1 row and dim columns and each element is uniformly distributed in the interval (0,1).

[0081] C2034. After resetting, update the global optimal solution and clear the stall counter. .

[0082] In step S3, the specific steps for optimizing the PID controller parameters using the improved barber optimization algorithm are as follows:

[0083] Step 1: Construct and define a PID control system model for dissolved oxygen in fishery waters. The controlled object is the aeration system for aquaculture. The specific transfer function model is as follows: , where s is a complex variable.

[0084] This study employs a system decomposition and synthesis approach, breaking down the entire aeration control system into three main components: actuators, controlled processes, and sensors, and modeling each component separately. The overall transfer function of the system is... It is the product of the transfer functions of each component, that is... .

[0085] Among them, for the actuator model The actuator is an aeration pump that regulates dissolved oxygen in the water. When the controller issues a power adjustment command, the pump motor experiences a brief delay in response due to its electrical and mechanical inertia, and cannot complete the adjustment instantaneously. This dynamic process can be accurately described by a first-order hysteresis loop, and its transfer function model is as follows: By setting the actuator gain (Normalized), time constant Seconds, according to the response characteristics of the aeration pump, the actuator model is as follows: .

[0086] Among them, for the controlled process model The controlled process is the aquaculture water itself, and the change in its dissolved oxygen concentration is a complex dynamic equilibrium process, mainly affected by two factors: the amount of oxygen injected by the aeration pump and the amount of oxygen consumed by biological respiration and organic matter decomposition in the water. Furthermore, the diffusion of oxygen from the aeration point to the sensor measurement point takes time, which constitutes the pure delay of the system. This process can be described by a first-order lag element connected in series with a pure delay element, and the transfer function model is as follows: Based on the principle of water mass balance, the process gain is estimated and set. This reflects the relationship between aeration rate and dissolved oxygen concentration under steady-state conditions; a time constant is set. The second reflects the large time lag characteristic of slow changes in dissolved oxygen concentration in large water bodies; the pure delay time is set. The second represents the delay in oxygen transport in the water body, and the controlled process model is as follows: .

[0087] Among them, the dissolved oxygen sensor model The sensor is a dissolved oxygen probe used to measure the real-time dissolved oxygen concentration. However, due to the time required for chemical or optical reactions and diffusion within the probe, the measured value lags behind the true value. The transfer function model is as follows: By setting the sensor gain time constant Seconds, therefore the sensor model is: .

[0088] In summary, by cascading and multiplying the transfer functions of the three components mentioned above, we obtain the overall open-loop transfer function of the system: To facilitate calculations in a simulation environment, the pure delay term in the model needs to be adjusted. For rational function approximation, the first-order Pader approximation is used, and its expression is: Substituting the Ped approximation into the overall model and performing polynomial expansion and simplification, we finally obtain a fourth-order transfer function model that can fully describe the dynamic characteristics of the controlled object. : This can relatively accurately reflect the complex characteristics of aquaculture oxygenation systems, such as slow response, large time lag, and the existence of initial reverse response.

[0089] Step 2: Configure the core parameters and fitness function of the improved barber optimization algorithm IBaOA, where the population size N=50; problem dimension dim=3; maximum number of iterations max_iter=100; lower bound of search space lb=[0.5,0.001,0.1]; upper bound of search space ub=[15.0,0.5,10.0]; hybrid guidance ratio parameter Cr=0.2; and stagnation tolerance upper bound. =15, Reset ratio =0.15; fitness function The cost function to be minimized is expressed as:

[0090] ,

[0091] The formula for multiplying time by the integral of absolute error ITAE is as follows: This is used to comprehensively evaluate the system's speed and stability. The simulation is performed on a step response with a setpoint of 2.0, and the total simulation time is 6000 seconds; an overshoot penalty term is included. The formula is When overshoot Exceeding the limit When this penalty is applied, the penalty weight is... The value is 10000; steady-state time penalty term. : When the steady-state time Exceeding the limit This penalty is applied at the specified time. Penalty weight. The value is 10000; weighting coefficient: The value is set to 1.0; and before calculating the fitness, the stability of the closed-loop system is checked. If it is unstable, a huge cost value is returned directly, guiding the algorithm to abandon the set of parameters.

[0092] Step 3: Execute the improved barber optimization algorithm to optimize PID parameters. After the algorithm starts, it first searches within the preset search space. Inside, according to the formula Fifty individuals are randomly generated as the initial population, and their fitness is evaluated. In the subsequent 100 iterations, the position vector of each individual is updated through the following exploration, exploitation, and stagnation reset phases.

[0093] (1) Exploration phase: For each individual First, generate a Random numbers in the interval and mixed guided scaling parameters Compare with 0.2; if the random number is less than 0.2. Then a hybrid guidance strategy is adopted, through the formula Generate a hybrid global optimal solution and random solutions A brand new guided solution Otherwise, a conventional strategy is used, selecting a pilot solution from among the more fit individuals in the population. Determine the guiding solution Then, update the formula based on exploration. Calculate the new position, where for Random vectors in an interval Given an integer vector with random values ​​of 1 or 2; calculate the fitness of the new position, and accept the update if it is better than the current one.

[0094] (2) Utilization phase: The position of an individual is updated using an elite-guided precision pruning strategy; the new position is determined by the current position. A path toward the global optimum Learning items And a random perturbation term that decays with the number of iterations. The sum of the three is, i.e. The formula for calculating the learning term is: Among them, learning factor Take 1.8, for A random vector within an interval; the formula for calculating the random disturbance term is: Among them, learning factor Take 0.5, This represents the current iteration number. The maximum number of iterations is 100. for A random vector within an interval; calculate the fitness of the new position, and if it is better than the current one, accept the update again.

[0095] (3) Stagnation Reset Mechanism: After each iteration, check whether the fitness value of the global optimum has improved. If it has not improved for 15 consecutive iterations, that is... If this happens, the reset mechanism will be activated, removing the bottom 15% of individuals in the population, i.e. According to the formula Re-initialize randomly to increase population diversity and help the algorithm escape local optima.

[0096] Step 4: In Matlab, the duration is set to 6000 seconds, and the step signal amplitude is 2.0. After 100 iterations, the improved barber optimization algorithm IBaOA converges to an optimal fitness value of 9823.7864, significantly lower than BaOA's 12280.0780. Output the optimal fitness value and the corresponding optimal PID parameter combination. =2.1646、 =0.0010、 =10.0000.

[0097] Among them, the improved barber optimization algorithm IBaOA is compared with the barber optimization algorithm BaOA, such as Figure 4As shown, the improved barber optimization algorithm IBaOA converges to an optimal fitness value of 9823.7864, significantly lower than the 12280.0780 of the barber optimization algorithm BaOA, indicating higher optimization accuracy and finding a solution closer to the optimum. Furthermore, the fitness convergence curve shows that the improved barber optimization algorithm IBaOA decreases rapidly in the early stages of iteration, quickly converging to near the final optimum in only about 11 iterations. In contrast, while the barber optimization algorithm BaOA also decreases rapidly in the early stages, it falls into a local optimum after about 15 iterations, and its fitness value shows no further significant improvement. The rapid decline of the improved barber optimization algorithm IBaOA curve near the 8th-10th iteration likely reflects the effectiveness of the stagnation reset and hybrid guided exploration mechanisms, successfully helping the algorithm escape early traps and find a better solution space.

[0098] like Figure 5 As shown, the proportional coefficients finally found by the two algorithms The values ​​are very close, all converging around 2.16. The Barber Optimization Algorithm BaOA is 2.1624, and the improved Barber Optimization Algorithm IBaOA is 2.1646, indicating that for this controlled object, the optimal proportional gain is within a relatively clear range.

[0099] like Figure 6 As shown, the integral coefficients finally found by the two algorithms All converged to a minimum value of 0.0010, which is represented by a nearly zero straight line in the figure. This indicates that for the dissolved oxygen PID control system in fishery water management, the integral action needs to be very weak in order to avoid integral saturation and oscillation.

[0100] like Figure 7 As shown, the differential coefficients finally found by the two algorithms The Barber Optimization Algorithm BaOA converged to a value of 5.4507, while the improved Barber Optimization Algorithm IBaOA found a value much larger than the former, reaching the upper bound of 10.0000. This indicates that the Barber Optimization Algorithm BaOA is relatively conservative and converged to a suboptimal value earlier.

[0101] like Figure 8 As shown, the PID controller with parameters optimized using the Barber Optimization Algorithm (BaOA) produces a large overshoot and a larger oscillation in the response curve, while the PID controller with parameters optimized using the improved Barber Optimization Algorithm (IBaOA) produces a smaller overshoot, a smaller oscillation in the response curve, and can stabilize near the set value more quickly, exhibiting better damping characteristics and robustness.

[0102] like Figure 2 As shown, the specific process of applying the optimized parameters to the control system in step S4 is as follows:

[0103] S401. When the preset periodic tuning time is met or a significant drop in system performance is detected, the improved barber optimization algorithm module is automatically triggered to start the parameter optimization process.

[0104] S402, Improved Barber Optimization Algorithm Module Based on Fitness Function Evaluate individual fitness and iteratively update the population through exploration, exploitation, and stagnation reset strategies;

[0105] S403. When the algorithm reaches the termination condition, extract the three component values ​​sequentially from the final output global optimal individual position vector, and parse them into proportional gain parameters respectively. Integral coefficient and differential coefficients The specific value;

[0106] S404. Through the parameter configuration interface of the dissolved oxygen PID control system for fishery water management, the parsed... , and The numerical value is written to the corresponding parameter register of the dissolved oxygen PID controller module so that the controller can immediately use the new parameters for optimized control.

Claims

1. A method of fishery water body dissolved oxygen control optimization, characterized by: Comprise the following steps: S1, construct fishery water body management dissolved oxygen PID control system, including dissolved oxygen error calculation module, dissolved oxygen PID controller module, improved barber optimization algorithm module, dissolved oxygen regulation module, dissolved oxygen monitoring module; S2, introduce improved barber optimization algorithm, the specific improvement strategy is: S201、In the exploration stage, introduce a mixed guidance proportion parameter Cr, and generate a guidance solution for guiding the position update of each current individual The generation strategy of the guidance solution is determined by the mixed guidance proportion parameter , specifically including: When the random number randomly taken from the interval [0, 1] is less than Cr, a hybrid guiding strategy is adopted to generate a guiding solution The specific formula is: ， wherein is the generated guiding solution, is a random weight in the interval [0, 1], is the current global optimum solution, is a randomly selected individual from the population; When the random number randomly taken from the interval [0, 1] is not less than , the guiding solution is generated along with the conventional strategy, and a solution with a better fitness value than the current individual is randomly selected from the entire population as the guiding solution . If there is no better solution, a solution is randomly selected from the entire population except the individual itself as the guiding solution . S202, in the utilization stage, the accurate pruning strategy of elite guidance is adopted, and a new position updating formula is specifically adopted: ， wherein is a learning step size towards the global optimal solution, is a random perturbation term, is the current position; The calculation formula of the learning step to the global optimal solution is: ， wherein is an elite guide learning factor, is a random vector in the interval [0, 1], is the current position, is the global optimum; Random perturbation term The calculation formula is: ， wherein is a random disturbance learning factor, is the current iteration number, is the maximum iteration number, is a random vector in the interval [0, 1], lb is the lower limit of the search space, and ub is the upper limit of the search space. S203, introduce stagnation detection and population reset mechanism, when the global optimal solution is not updated for a preset number of times, reset the population according to a preset reset proportion determine the number of poor individuals to be reset, and reset the determined number of poor individuals​ S3, using improved barber optimization algorithm for fishery water management dissolved oxygen PID control system of dissolved oxygen PID control parameter setting, through the optimization of the best control parameters 、 、 ; S4, set the three optimal control parameters obtained by using the improved barber optimization algorithm to optimize the dissolved oxygen regulation control effect.

2. A method of fishery water body dissolved oxygen control optimization according to claim 1, characterized in that: In the fishery water body management dissolved oxygen PID control system constructed in step S1, the actual dissolved oxygen value is collected by the dissolved oxygen monitoring module and transmitted to the dissolved oxygen error calculation module, the target dissolved oxygen is set and received by the dissolved oxygen error calculation module, and after the error calculation of the target dissolved oxygen and the actual dissolved oxygen, the real-time error signal is output to the dissolved oxygen PID controller module, the internal parameters of the PID controller module are continuously optimized by the improved barber optimization algorithm IBaOA module to optimize the control performance, and the optimized PID controller module calculates the control amount according to the error signal, and then outputs it to the dissolved oxygen adjusting module to perform the adjusting operation of the dissolved oxygen concentration. 、 、 ​ 3. A method of fishery water body dissolved oxygen control optimization according to claim 1, characterized in that: In step S203, the stagnation detection and population reset mechanism is added, comprising the following steps: C2031, after each iteration, check the global optimum solution update status, if the optimal fitness value is not improved continuously, then the stagnation counter increment 1; C2032、When the reset mechanism is triggered, the first individuals are selected, in descending order of fitness value, where is the population size. C2033, according to the reset formula , re-initialize the poor individual position and evaluate the fitness, where lb is the lower limit of the search space, ub is the upper limit of the search space, dim is the problem dimension and takes the value of 3, and the rand function is used to generate a random vector of 1 row and dim columns, and each element is uniformly distributed in the interval (0, 1); C2034. Update global optimum solution after reset and clear stall counter .

4. A method of fishery water body dissolved oxygen control optimization according to claim 1, characterized in that: In step S3, the specific steps of optimizing the PID controller parameters by using the improved barber optimization algorithm are: S301, initialize the core parameters of the improved barber optimization algorithm, including the population size , problem dimension , maximum number of iterations , lower limit of search space and upper limit , mixed guidance proportion parameter , upper limit of stagnation tolerance , reset proportion , and randomly generate initial population positions in three-dimensional space; S302, a mapping relationship between an algorithm solution space and PID controller parameters is established, and three components of an individual position vector respectively correspond to proportional coefficients, integral coefficients and differential coefficients of the PID controller ​​​ S303、Define fitness function for evaluating PID control performance The mathematical expression is: ， wherein is the time multiplied by the absolute error integral, is the overshoot of the system response, is the overshoot penalty term, is the overshoot upper limit, is the settling time of the system response, is the settling time penalty term, , , is the weight coefficient, is the settling time of the response, is the settling time upper limit of the response; S304, execute the algorithm core update mechanism, loop iteration, in each iteration, update and utilize the exploration stage for each individual in turn, and execute the stagnation detection and population reset mechanism at the end of each iteration; S305、when the number of iterations reaches the optimization is terminated, and the global optimal individual position vector is output as the optimal parameter combination of the PID controller 、 、 .

5. A method of fishery water body dissolved oxygen control optimization according to claim 1, characterized in that: In step S4, the specific process of applying the optimized parameters to the control system is: S401, when the preset periodic setting opportunity is met or the system performance is detected to be significantly reduced, the improved barber optimization algorithm module is automatically triggered to start the parameter optimization process. S402、the improved barber optimization algorithm module is based on fitness function evaluate the fitness of individuals and iteratively update the population through exploration, exploitation, and stagnation reset strategies; S403、When the algorithm reaches the termination condition, extract the three component values in order from the final output global optimal individual position vector, and respectively analyze into specific values of proportional gain parameter , integral coefficient and differential coefficient ​ S404, write the parsed values into the corresponding parameter registers of the dissolved oxygen PID controller module through the system parameter configuration interface, so that the controller immediately adopts the new parameters for optimization control. 、 and values into the corresponding parameter registers of the dissolved oxygen PID controller module to enable the controller to immediately adopt the new parameters for optimization control.

Images