Colla corii asini liquid temperature control optimization method for colla corii asini processing

Abstract

The invention discloses a gelatin solution temperature control optimization method for donkey-hide gelatin processing, and belongs to the technical field of swarm intelligence optimization. Comprising the steps that a gelatin solution temperature closed-loop PID control system in the donkey-hide gelatin decocting process is constructed, an improved side hand spider turning optimization algorithm is configured to serve as a parameter setting core, the self-adaptive setting process of PID control parameters is executed, and a global optimal control parameter vector obtained through optimization is mapped to a donkey-hide gelatin solution temperature PID controller module. The method focuses on the PID parameter setting problem of donkey-hide gelatin liquid temperature control, introduces the improved side hand turning spider optimization algorithm as a swarm intelligence algorithm to set the donkey-hide gelatin liquid temperature PID control parameters, and can effectively overcome the defect that the side hand turning spider optimization algorithm is easy to fall into local optimum. And the temperature of the donkey-hide gelatin liquid can be controlled more excellently.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of swarm intelligence optimization, and particularly relates to a glue liquid temperature control optimization method for processing donkey-hide gelatin. BACKGROUND

[0002] As a core link in the production process of donkey-hide gelatin, the temperature control of glue liquid directly determines the physicochemical properties and efficacy stability of the glue liquid. In the boiling process, the accurate control of the glue liquid temperature is the cornerstone of guaranteeing product quality. However, the glue liquid has the characteristics of large inertia and large time lag in the heating process, and the viscosity of the glue liquid changes nonlinearly with the evaporation of water, which easily causes temperature fluctuations. The excessively high temperature can cause protein denaturation, damage effective components, and even cause production safety hazards such as boiling over. The current general solution is to use the traditional PID control strategy to adjust the heating power through temperature feedback. Although the PID control structure is simple, the conventional PID controller often faces the problems of parameter tuning difficulty, large response lag, low control precision and the like when facing the complex time-varying characteristics such as viscosity mutation in the boiling process of donkey-hide gelatin, and it is difficult to meet the constant temperature demand of high-quality donkey-hide gelatin production. SUMMARY

[0003] In order to overcome the technical problems described in the background, the application provides a glue liquid temperature control optimization method for processing donkey-hide gelatin. The PID parameter tuning problem of the temperature control of donkey-hide gelatin glue liquid is focused on, and an improved side handstand spider optimization algorithm as a swarm intelligence algorithm is introduced to tune the PID control parameters of the temperature of donkey-hide gelatin glue liquid. The improved side handstand spider optimization algorithm is superior to the side handstand spider optimization algorithm in terms of convergence speed, optimization accuracy, overshoot suppression and regulation speed of the controlled system. The improved side handstand spider optimization algorithm can effectively overcome the defect that the side handstand spider optimization algorithm is easy to fall into local optimum, and can make the temperature of donkey-hide gelatin glue liquid obtain a relatively more excellent control effect.

[0004] The technical scheme of the application is as follows: a glue liquid temperature control optimization method for processing donkey-hide gelatin, comprising the following steps: S1, constructing a glue liquid temperature closed-loop PID control system for the boiling process of donkey-hide gelatin, including a glue liquid temperature error calculation module, a glue liquid temperature PID controller module, an improved side handstand spider optimization algorithm module, a glue liquid temperature adjustment execution module and a glue liquid temperature real-time monitoring module; S2, configuring the improved side handstand spider optimization algorithm as a parameter tuning core, and the specific algorithm improvement strategy is configured as: S21. For the global exploration stage where individual spiders are stuck in deep stagnation, a spider web vibration triangulation strategy is embedded. The individual position is reconstructed by introducing the mutation mechanism of the differential evolution operator. The vector difference information between the population is used to guide the stagnant individual to perform directional jumps, so as to enhance the algorithm's ability to escape local extreme value traps in complex fitness landscapes. S22. For spiders in a normal exploration state, the Levy flight predation and attack strategy is integrated. The long-tailed distribution random step size generated by the Mantegna algorithm is used to replace the linear translation component, and a translation guidance term with non-uniform variable step size characteristics is constructed to improve the global traversal efficiency and coverage of the algorithm in the broad PID parameter solution space. S23. For the local development stage where individual spiders have sufficient energy, a chaotic silk-spinning micro-weaving strategy is implanted. The ergodicity and randomness of the Tent chaotic mapping are used to generate chaotic perturbation factors, and the current solution is refined for search and neighborhood fine-tuning to improve the algorithm's accuracy and convergence speed in the global optimal solution region. S3. Execute the adaptive tuning process for the PID control parameters, and use the improved side-flipping spider optimization algorithm to iteratively optimize the control parameter space of the adhesive temperature PID control system to obtain the optimal control parameter vector that balances system response speed and stability. , , ; S4. Map the globally optimal control parameter vector obtained through optimization to the PID controller module for the temperature of the donkey-hide gelatin solution, update the internal parameters of the controller, thereby achieving precise adjustment and anti-disturbance control optimization of the temperature of the gelatin solution during the donkey-hide gelatin cooking process.

[0005] Furthermore, in the PID control system for the temperature of the gelatin liquid in the gelatin processing constructed in step S1, historical operating data is collected through the gelatin liquid temperature monitoring module, and a transfer function model of the gelatin liquid temperature control system is established. The improved side-hand-turning spider optimization algorithm module performs iterative optimization based on the transfer function model, and calculates the PID controller parameters in a virtual simulation environment. After the iteration termination condition is met, the optimal global control parameters obtained by optimization are output and configured into the adhesive temperature PID controller module. The adhesive temperature PID controller module calculates the precise control quantity based on the real-time temperature error signal and the configured optimal control parameters, and outputs the control quantity to the adhesive temperature adjustment module to drive the actuator to accurately adjust the adhesive temperature.

[0006] Furthermore, the spiderweb vibration triangulation localization strategy in step S21 includes the following steps: S211. In global exploration mode, first check the stagnation counter of the current spider individual. Does it exceed the threshold of 5? S212, if the following condition is satisfied , the position vector of the third individual is denoted as , and ; S213, based on the principle of differential evolution, the difference between the last two individuals is weighted by the scaling factor , and is superimposed on the first individual to generate a new position, the calculation formula is: , where is the generated new position vector, is the position vector of the three different individuals randomly selected from the current population, is the scaling factor of differential evolution.

[0007] Further, the Levy flight predation strike strategy in step S22 includes the following steps: S221, when the spider individual enters the directional somersault state, that is, when , first, based on the Mantegna algorithm, a random step vector that obeys the Levy distribution is generated using the Levy index ; S222, the rotation component is calculated , where is the rotation component, is the strength of the side somersault, is a random number between 0 and 1, and are the upper bound vector and the lower bound vector of the search space, respectively, is the rotation range coefficient; S223, the translation guide item is calculated using the Levy step vector , the calculation formula is: , where is the translation guide item, is the Levy flight step scaling coefficient, is the Levy flight random step vector, is the current global best position, is the current individual position, and the symbol indicates that the elements of the corresponding dimensions of the vectors are multiplied; S224, the rotation component and the Levy translation component are superimposed on the current position to complete the position update.

[0008] Further, the chaotic spinning micro-weaving strategy in step S23 comprises the following steps: S231, when the energy level of the spider individual is less than the energy threshold, enter the chaotic local search mode, initialize the chaotic variable , and set the internal search iteration number ; S232, calculate the current chaotic search radius adjusted dynamically according to the development factor , and the calculation formula is: , wherein is the chaotic search radius, is the development factor increasing with the increase of the iteration number, and are the upper bound vector and the lower bound vector of the search space respectively, is the basic scaling coefficient of the search radius; S233, in the times loop, the chaotic variable is iteratively updated by using the Tent mapping formula, and the disturbance factor with more uniform distribution is generated by using the ergodic property thereof; S234, generate the candidate solution based on the current local optimal position , and the calculation formula is: ; If the fitness of the candidate solution is better, the local position is immediately updated.

[0009] Further, the mathematical expression of the optimization objective function of the EJ glue liquid temperature PID control system is: , wherein represents the total simulation time length, is the time, is the real-time deviation of the set temperature and the actual glue liquid temperature, is the actual overshoot of the system step response, is the maximum overshoot threshold allowed, is the penalty weight of the constraint violation.

[0010] The present application has the beneficial effects due to the adoption of the above-mentioned technology: ​1. The present application focuses on the problem of the PID parameter setting of the donkey-hide gelatin glue temperature control, introduces the improved side handstand spider optimization algorithm as a group intelligence algorithm to set the PID control parameter of the donkey-hide gelatin glue temperature, the improved side handstand spider optimization algorithm is superior to the side handstand spider optimization algorithm in convergence speed, optimization accuracy and overshoot suppression and regulation speed of the controlled system, the improved side handstand spider optimization algorithm can effectively overcome the defect that the side handstand spider optimization algorithm is easy to fall into local optimum, and can make the donkey-hide gelatin glue temperature obtain a relatively more excellent control effect; 2. The present application configures the spider web vibration triangular positioning strategy in the improved side handstand spider optimization algorithm module, reconstructs the position of the spider individual in a deep stagnation state by using the mutation mechanism of the differential evolution operator, guides the stagnation individual to perform directional jumping by using the vector difference information between different individuals in the population, thereby destroying the balance state of the local extreme point, so that the algorithm can force the solution vector to escape from the local optimal trap in the multi-peak PID parameter fitness landscape, and ensure the continuous evolution ability of the algorithm in the complex solution space; 3. The present application integrates the Levi flight hunting strategy, uses the random step length generated by the Mantegna algorithm to replace the linear translation component in the original algorithm, the Levi flight has the non-uniform variable step length feature that long and short steps appear alternately, so that the search agent can perform long-distance cross-region jumping and short-distance fine search in the wide PID parameter solution space, expand the traversal range of the algorithm in the three-dimensional solution space composed of proportional coefficient, integral coefficient and differential coefficient, and reduce the probability that the algorithm misses the region where the global optimal solution is located; 4. The present application implants the chaotic silk spinning micro-weaving strategy, uses the ergodicity and randomness of Tent chaotic mapping to generate chaotic disturbance factors to replace the single cosine rolling mechanism in the original algorithm, applies chaotic disturbance to the current solution in the local development stage, so that the algorithm can perform high-density traversal search in the neighborhood of the candidate solution, and improve the mining accuracy of the global optimal control parameter of the algorithm, thereby reducing the steady-state error of the glue temperature control system. BRIEF DESCRIPTION OF DRAWINGS

[0011] Figure 1 It is the flowchart of the glue temperature control optimization method of the present application.

[0012] Figure 2 It is the optimal fitness convergence curve comparison diagram of the side handstand spider optimization algorithm and the improved side handstand spider optimization algorithm of the present application.

[0013] Figure 3 It is the step response curve comparison diagram of the side handstand spider optimization algorithm and the improved side handstand spider optimization algorithm of the present application.

[0014] Figure 4 This is a comparison chart of the PID parameter search trajectories of the side-hand-turning spider optimization algorithm of the present invention and the improved side-hand-turning spider optimization algorithm. Detailed Implementation

[0015] Example 1: As Figure 1 As shown, the present invention provides an optimization method for controlling the temperature of the gelatin solution in the processing of donkey-hide gelatin, comprising the following steps: S1. Construct a closed-loop PID control system for the gelatin liquid temperature during the gelatin boiling process, including a gelatin liquid temperature error calculation module, a gelatin liquid temperature PID controller module, an improved side-hand-turning spider optimization algorithm module, a gelatin liquid temperature regulation execution module, and a gelatin liquid temperature real-time monitoring module. The gelatin liquid temperature monitoring module collects historical operating data and establishes a transfer function model for the gelatin liquid temperature control system. The improved side-hand-turning spider optimization algorithm module performs iterative optimization based on the transfer function model, calculating the PID controller parameters in a virtual simulation environment. After the iteration termination condition is met, the best global control parameters obtained by optimization are output and configured into the adhesive temperature PID controller module. The adhesive temperature PID controller module calculates the precise control quantity based on the real-time temperature error signal and the configured best control parameters, and outputs the control quantity to the adhesive temperature adjustment module to drive the actuator to precisely adjust the adhesive temperature. S2. Configure the improved side-hand-turning spider optimization algorithm as the core of parameter tuning. The specific algorithm improvement strategy is configured as follows: S21. For the global exploration stage where individual spiders are stuck in deep stagnation, a spider web vibration triangulation strategy is embedded. The individual position is reconstructed by introducing the mutation mechanism of the differential evolution operator. The vector difference information between the population is used to guide the stagnant individual to perform directional jumps, so as to enhance the algorithm's ability to escape local extreme value traps in complex fitness landscapes. S22. For spiders in a normal exploration state, the Levy flight predation and attack strategy is integrated. The long-tailed distribution random step size generated by the Mantegna algorithm is used to replace the linear translation component, and a translation guidance term with non-uniform variable step size characteristics is constructed to improve the global traversal efficiency and coverage of the algorithm in the broad PID parameter solution space. S23. For the local development stage where individual spiders have sufficient energy, a chaotic silk-spinning micro-weaving strategy is implanted. The ergodicity and randomness of the Tent chaotic mapping are used to generate chaotic perturbation factors, and the current solution is refined for search and neighborhood fine-tuning to improve the algorithm's accuracy and convergence speed in the global optimal solution region. S3. Execute the adaptive tuning process for the PID control parameters, and use the improved side-flipping spider optimization algorithm to iteratively optimize the control parameter space of the adhesive temperature PID control system to obtain the optimal control parameter vector that balances system response speed and stability. , , ; S4, mapping the global optimal control parameter vector obtained by optimization to the colla glue liquid temperature PID controller module, updating the internal parameters of the controller, so as to realize precise adjustment and anti-disturbance control optimization of the glue liquid temperature in the colla boiling process.

[0016] Wherein, the control parameter space of the glue liquid temperature PID control system is iteratively optimized by using the improved side somersault spider optimization algorithm in step S3 to obtain the optimal PID control parameter vector of the colla glue liquid temperature control system , specifically comprising the following steps: step1, initialization of PID parameter optimization of colla glue liquid temperature PID control system; step11, the mathematical expression of the optimization objective function of colla glue liquid temperature PID control system is: , In the formula , the total simulation time is 2500 seconds, is the time, is the real-time deviation of the set temperature and the actual glue liquid temperature, is the actual overshoot of the system step response, is the maximum overshoot threshold allowed and the value is 2.0, is the penalty weight of constraint violation and the value is 100, wherein the objective function The smaller the function value is, the better the control performance is; At the same time, the population size and the maximum number of iterations are set; step12, according to the high viscosity and large time delay characteristics of the colla glue boiling process and combined with the static gain characteristics of the controlled object, the search space upper bound vector of the PID control parameter is set as = and the search space lower bound vector is = ; step13, initialization of key hyperparameters of improved algorithm; The differential evolution scaling factor is set to 0.5, the Levy flight index is set to 1.5, the number of iterations of chaotic local search is set to 5, the side somersault intensity coefficient is set to 0.6, the rotation range coefficient is set to 0.1, and the Levy flight step scaling coefficient is 0.01; step 14, randomly initialize the population in the search space spider population position of group PID control parameters , set the initial energy level of each spider to 1.0, set the stagnation counter to 0; step 15, calculate the fitness value of each individual in the initial population, mark the individual with the best fitness as the initial global optimal PID parameter scheme .

[0017] step 2, spider population iterative optimization stage based on hybrid strategy driven; step 21, start the main loop; current iteration number from 1 to , calculate the exploration factor and development factor that changes over time , , wherein represents the current iteration number, represents the maximum iteration number; step 22, traverse each spider individual in the population , calculate the threshold value that triggers behavior transition , wherein represents the current energy level of the th individual; step 23, generate a random number , if is less than , enter global exploration mode and execute step 24, otherwise enter local development mode and execute step 25; step 24, execute the improved side handstand global exploration strategy; step 241, check the stagnation counter of the current individual , if is greater than 5, execute the cobweb vibration triangular positioning strategy; , wherein is the new position, for three mutually different individual positions randomly selected from the population, is a differential scaling factor; step 242, if is less than or equal to 5, a Levy flight predatory strike strategy is executed; First, based on the Mantegna algorithm, for each dimension of the PID parameter respectively, the Levy flight step component is calculated independently, and the calculation formula is , , In the formula, is the gamma function, the value is 1.5, and are the random numbers generated independently according to the normal distribution for the dimension ; The three-dimensional column vector is composed of each component , and then the individual position is updated, at which time the formula of the new position is: , In the formula, is the current individual position, is a random number, and are the lower and upper bound vectors of the search space corresponding to the dimension, and the multiplication operation is performed dimension by dimension, respectively, the coefficients set in the foregoing, and the symbol represents the multiplication of the elements of the corresponding dimension of the vector; step 243, boundary constraint processing is performed on the generated new position to ensure that is located within ; step 25, an improved rolling local development strategy is executed; step 251, the energy level of the current individual is checked, and if is less than 0.2, an energy recovery hibernation adjustment strategy is executed, and the formula of the new position is: , In the formula, is the global best position; step 252, if is greater than or equal to 0.2, a chaotic silk spinning micro-weaving strategy is executed; First, the chaotic variable is randomly initialized, Belongs to the interval Ensure initial values The values ​​are not equal to 0.25, 0.5, or 0.75 to avoid a fixed point; Subsequently The iteration formula for the Tent mapping is as follows: (The text is incomplete and requires further context.) , In the formula For the first The values ​​of the chaotic variables in the next iteration; Candidate solutions are generated based on the updated chaotic variables, and the calculation formula is as follows: , In the formula This is the current local optimum. The radius of the chaotic search; in In the formula The development factor is the factor that increases with the number of iterations. and Let be the upper and lower bound vectors of the search space, respectively. This is the base scaling factor for the search radius, and its value is 0.05. if Better than Then update the local optimum, and the final output is ; Step 26: Evaluate the new location fitness value ; if Superior to the old fitness If the new solution is accepted, energy is increased and the stagnation count is reset. The calculation formula is as follows: , ; At the same time if If the fitness is better than the global best, then update. ; Step 27, if Not better than If the old solution is retained, the energy is reduced and the stagnation count is accumulated. The calculation formula is: , ; Step 28: If an individual has performed step 251 in the current iteration, then after completing the evaluation and state update, force the individual's energy to be reset to [value missing]. .

[0018] step3, optimal temperature control parameter output stage; step31, judge whether the maximum iteration number is reached , if not, return to step21 to continue the next round of iteration; step32, if the termination condition is reached, output the global optimal position , That is, the optimal PID control parameter vector of the Ejiao glue liquid temperature control system .

[0019] Among them, based on the high viscosity and thermal inertia characteristics of the Ejiao glue liquid during the boiling process, the transfer function of the Ejiao glue liquid temperature PID control system of the controlled object is constructed The mathematical expression is: , In the formula is the Laplace operator, after system identification, represents the system static gain and takes the value of 50.0, represents the system time constant and takes the value of 500.0, represents the system pure lag time and takes the value of 30.0.

[0020] And correspondingly, the control law of Ejiao glue liquid temperature PID controller of Ejiao glue liquid temperature PID control system , adopts parallel structure, and its time domain expression is: , In the formula is the deviation of the set temperature and the actual output temperature, is the proportional coefficient, is the integral coefficient, is the differential coefficient; these three parameters are the decision variable vector optimized by the algorithm .

[0021] In order to verify the difference of the side somersault spider optimization algorithm and the improved side somersault spider optimization algorithm in the PID parameter optimization performance of Ejiao glue liquid temperature control system, the project code is realized in Matlab environment and the general running parameters of the optimization algorithm are set. The population size takes the value of 50, the maximum iteration number takes the value of 200, the random number seed rng is configured, and the value is 1 to ensure the reproducibility of the comparative experiment, and the side somersault spider optimization algorithm and the improved side somersault spider optimization algorithm are run respectively in the same experimental environment and the optimal fitness convergence curve, the optimal PID parameter solution and the corresponding step response curve are recorded; As Figure 2The best fitness convergence curve contrast chart shows that the final fitness value of the improved side somersault spider optimization algorithm is 1016.9036, which is significantly lower than 4336.4848 of the side somersault spider optimization algorithm. The lower fitness value directly proves that the improved side somersault spider optimization algorithm has higher optimization accuracy. The improved side somersault spider optimization algorithm shows a sharp downward trend in the initial iteration of about 20 generations, rapidly approaches the global optimal solution, and embodies strong evolutionary pressure and exploration ability. In contrast, the side somersault spider optimization algorithm converges slowly and finally falls into a local optimal trap, which is premature convergence and fails to effectively explore better parameter space. As Figure 3 shown in the step response curve contrast chart, the system response of the improved side somersault spider optimization algorithm after tuning almost realizes the "quasi-no overshoot" feature, and the waveform is smooth. In contrast, the response curve corresponding to the side somersault spider optimization algorithm has an obvious overshoot of more than , accompanied by obvious oscillation phenomenon, indicating that the parameter combination cannot effectively balance the damping characteristics of the system. Moreover, the improved side somersault spider optimization algorithm system enters the steady state range within about , but the adjustment time of the side somersault spider optimization algorithm lags far behind due to severe oscillation, and the dynamic recovery performance of the system is poor. As Figure 4 shown in the PID parameter search trajectory contrast chart, in the optimization process, three dimensions exhibit good collaborative convergence characteristics under the drive of the improved side somersault spider optimization algorithm. In the late iteration, the parameter search trajectory of the improved side somersault spider optimization algorithm quickly tends to be a straight line, verifying that the algorithm has strong local fine-tuning ability in the later period and ensuring the numerical stability of the final output result. In summary, the improved side somersault spider optimization algorithm surpasses the side somersault spider optimization algorithm in terms of convergence speed, optimization accuracy, and overshoot suppression and regulation speed of the controlled system. The improved side somersault spider optimization algorithm can effectively overcome the defect of the side somersault spider optimization algorithm that is prone to falling into a local optimum. The optimal PID control parameter vector output is = [0.2463, 0.0005, 2.8423], which can achieve relatively more excellent control effect on the temperature of the colla glue solution.

Claims

1. A method for optimizing the temperature control of the gelatin solution used in the processing of donkey-hide gelatin, characterized in that: Includes the following steps: S1. Construct a closed-loop PID control system for the temperature of the gelatin solution during the gelatin boiling process, including a gelatin solution temperature error calculation module, a gelatin solution temperature PID controller module, an improved side-hand-turning spider optimization algorithm module, a gelatin solution temperature regulation execution module, and a gelatin solution temperature real-time monitoring module. S2. Configure the improved side-hand-turning spider optimization algorithm as the core of parameter tuning. The specific algorithm improvement strategy is configured as follows: S21. For the global exploration stage where individual spiders are stuck in deep stagnation, a spider web vibration triangulation strategy is embedded. The individual position is reconstructed by introducing the mutation mechanism of the differential evolution operator. The vector difference information between the population is used to guide the stagnant individual to perform directional jumps, so as to enhance the algorithm's ability to escape local extreme value traps in complex fitness landscapes. S22. For spiders in a normal exploration state, the Levy flight predation and attack strategy is integrated. The long-tailed distribution random step size generated by the Mantegna algorithm is used to replace the linear translation component, and a translation guidance term with non-uniform variable step size characteristics is constructed to improve the global traversal efficiency and coverage of the algorithm in the broad PID parameter solution space. S23. For the local development stage where individual spiders have sufficient energy, a chaotic silk-spinning micro-weaving strategy is implanted. The ergodicity and randomness of the Tent chaotic mapping are used to generate chaotic perturbation factors, and the current solution is refined for search and neighborhood fine-tuning to improve the algorithm's accuracy and convergence speed in the global optimal solution region. S3. Execute the adaptive tuning process for the PID control parameters, and use the improved side-flipping spider optimization algorithm to iteratively optimize the control parameter space of the adhesive temperature PID control system to obtain the optimal control parameter vector that balances system response speed and stability. , , ; S4. Map the globally optimal control parameter vector obtained through optimization to the PID controller module for the temperature of the donkey-hide gelatin solution, update the internal parameters of the controller, thereby achieving precise adjustment and anti-disturbance control optimization of the temperature of the gelatin solution during the donkey-hide gelatin cooking process.

2. The method for optimizing the temperature control of the gelatin solution in the processing of donkey-hide gelatin according to claim 1, characterized in that: In the PID control system for the temperature of the gelatin liquid in the donkey-hide gelatin processing constructed in step S1, historical operating data is collected through the gelatin liquid temperature monitoring module, and a transfer function model of the gelatin liquid temperature control system is established. The improved side-hand-turning spider optimization algorithm module performs iterative optimization based on the transfer function model and calculates the PID controller parameters in a virtual simulation environment. After the iteration termination condition is met, the optimal global control parameters obtained by optimization are output and configured into the adhesive temperature PID controller module. The adhesive temperature PID controller module calculates the precise control quantity based on the real-time temperature error signal and the configured optimal control parameters, and outputs the control quantity to the adhesive temperature adjustment module to drive the actuator to accurately adjust the adhesive temperature.

3. The method for optimizing the temperature control of the gelatin solution in the processing of donkey-hide gelatin according to claim 1, characterized in that, The spiderweb vibration triangulation localization strategy in step S21 includes the following steps: S211. In global exploration mode, first check the stagnation counter of the current spider individual. Does it exceed the threshold of 5? S212, if satisfied Three distinct individuals are randomly selected from the current population, and their position vectors are denoted as follows: , and ; S213. Based on the principle of differential evolution, using scaling factors The differences between the last two individuals are weighted and then added to the first individual to generate a new position. The calculation formula is as follows: ， In the formula For the generated new position vector, Let the vectors be the position vectors of three distinct individuals randomly selected from the current population. This is the scaling factor for differential evolution.

4. The method for optimizing the temperature control of the gelatin solution in the processing of donkey-hide gelatin according to claim 1, characterized in that, The Levi flight predator-prey attack strategy in step S22 includes the following steps: S221. When an individual spider enters a directional somersault state, that is, when... At that time, based on the Mantegna algorithm, the Lévy index was first used. Generate a random step-size vector that follows a Lévy distribution. ; S222, Calculate the rotation component , ， In the formula For rotational components, For the intensity of a side handspring, A random number between [0, 1] and Let be the upper and lower bound vectors of the search space, respectively. This is the rotation range coefficient; S223, Using Levy step size vector The translation guiding term is calculated using the following formula: ， In the formula As a translation guide item, This is the scaling factor for Levy's flight step size. Let Lévy's flight random step size vector be . This is the current global optimal position. The symbol represents the current individual's position. This represents element-wise multiplication of vectors along corresponding dimensions; S224. Superimpose the rotation component and the Levy translation component onto the current position. Up, complete the location update.

5. The method for optimizing the temperature control of the gelatin solution in the processing of donkey-hide gelatin according to claim 1, characterized in that, The chaotic micro-weaving strategy in step S23 includes the following steps: S231, When the energy level of an individual spider At this time, it enters the chaotic local search mode and initializes the chaotic variables. And set the number of internal search iterations. ; S232, Calculate the development factor Dynamically adjusted current chaotic search radius The calculation formula is: ， In the formula The radius of the chaotic search is... The development factor is the factor that increases with the number of iterations. and Let be the upper and lower bound vectors of the search space, respectively. This is the base scaling factor for the search radius; S233, in In the next iteration, the chaotic variables are iteratively updated using the Tent mapping formula. By utilizing its traversal properties, a more uniformly distributed perturbation factor can be generated. S234. Based on the current local optimal position The candidate solution is generated using the following formula: ； If a candidate solution has better fitness, the local position is updated immediately.

6. The method for optimizing the temperature control of the gelatin solution in the processing of donkey-hide gelatin according to claim 1, characterized in that, The optimization objective function of the PID control system for the temperature of donkey-hide gelatin solution The mathematical expression is: ， In the formula Indicates the total simulation duration. For time, To determine the real-time deviation between the set temperature and the actual adhesive temperature, This represents the actual overshoot of the system's step response. The maximum allowable overshoot threshold, The penalty weight for violating the constraint.

Images