Temperature control method and system based on IWOA optimization fuzzy PID

Abstract

The invention relates to the technical field of temperature control, and discloses a temperature control method and system based on IWOA optimization fuzzy PID, and the method comprises the steps: building a dual-channel data processing architecture according to an instruction generation channel and a parameter optimization channel; key indexes of the current dynamic behavior are extracted, differences of dimensions and orders of magnitude are eliminated, and standard temperature feature vectors are obtained; optimizing the standard temperature feature vector in a parameter space according to an iterative search strategy to obtain an optimized parameter set; the real-time temperature data sequence is reconfigured to a PID controller, the real-time temperature data sequence is coded, and a preliminary control instruction is generated; the amplitude is controlled to change suddenly, a stable control instruction is generated and smoothed, and a final control instruction is obtained; when the expected threshold value is not reached after the final control instruction is executed, the PID controller is regulated and controlled again; according to the invention, the efficiency of temperature control based on IWOA optimization fuzzy PID can be improved.

Description

Technical Field

[0001] This invention relates to the field of temperature control technology, and in particular to a temperature control method and system based on IWOA-optimized fuzzy PID. Background Technology

[0002] Traditional PID controllers rely heavily on manual experience or fixed rules for parameter configuration, making it difficult to adapt to the complex and ever-changing dynamic characteristics and external disturbances in temperature control systems. Their parameter adjustment lacks adaptability, which can lead to problems such as response lag and excessive overshoot during temperature control. They are unable to quickly and accurately track the preset temperature target, ultimately resulting in insufficient control accuracy and poor dynamic response performance.

[0003] Existing solutions based on intelligent algorithms to optimize fuzzy PID control have significant limitations in the parameter optimization stage. Some algorithms have slow convergence speeds and are prone to getting trapped in local optima, making it difficult to fully unleash the performance potential of the controller. At the same time, there is a lack of effective stability control mechanisms after the control commands are generated, and sudden changes in the command amplitude can easily cause system oscillations, further weakening the stability and reliability of temperature control. This makes it impossible to meet the actual needs of high-precision and high-stability temperature control in industrial production and other scenarios. Therefore, how to improve the efficiency of temperature control based on IWOA-optimized fuzzy PID control has become an urgent problem to be solved. Summary of the Invention

[0004] This invention provides a temperature control method and system based on IWOA-optimized fuzzy PID to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, this invention provides a temperature control method based on IWOA-optimized fuzzy PID, comprising:

[0006] S1. Generate a channel and a parameter optimization channel according to the instructions of the PID controller to construct a dual-channel data processing architecture for the temperature control system in the PID controller;

[0007] S2. Extract key indicators reflecting the current dynamic behavior of the PID controller from the real-time temperature data sequence collected by the PID controller, and eliminate the differences in dimensions and orders of magnitude in the key indicators to obtain the standard temperature feature vector of the PID controller.

[0008] S3. In the parameter optimization channel, the standard temperature feature vector is optimized in the parameter space of the PID controller according to the iterative search strategy to obtain the optimized parameter set of the PID controller.

[0009] S4. Through the instruction generation channel, the optimized parameter set is reconfigured into the PID controller, and the configured PID controller is used to encode the real-time temperature data sequence to generate the initial control instruction of the PID controller.

[0010] S5. Control the abrupt changes in the amplitude of the initial control command, generate a stable control command for the PID controller, and smooth the stable control command to obtain the final control command for the PID controller.

[0011] S6. When the expected threshold is not reached after the final control command is executed, the re-optimization condition of the parameter optimization channel is triggered to re-adjust the PID controller so as to obtain a PID controller that reaches the expected threshold.

[0012] In a preferred embodiment, the step of extracting key indicators reflecting the current dynamic behavior of the PID controller from the real-time temperature data sequence acquired from the PID controller, and eliminating differences in units and orders of magnitude among the key indicators to obtain the standard temperature feature vector of the PID controller, includes:

[0013] The real-time temperature data sequence is divided into multi-scale time windows, and the response speed index, stability index and steady-state performance index of the PID controller are extracted in different time windows respectively.

[0014] The response speed index, the stability index, and the steady-state performance index are fused to obtain a multi-dimensional dynamic behavior feature set of the PID controller;

[0015] Analyze the data distribution of the feature indicators in the multidimensional dynamic behavior feature set to obtain the numerical range and distribution characteristics of the feature indicators;

[0016] Based on the numerical range and the distribution characteristics, the feature index is mapped to a unified interval, and the data distribution pattern of the feature index is adjusted to obtain the standard feature index of the PID controller.

[0017] Correlation analysis is performed on the standard feature indicators, the key feature dimensions of the standard feature indicators are retained, and the key feature dimensions are quantified to obtain the standard temperature feature vector of the PID controller.

[0018] In a preferred embodiment, the step of fusing the response speed index, the stability index, and the steady-state performance index to obtain a multi-dimensional dynamic behavior feature set of the PID controller includes:

[0019] According to the degree of influence of the response speed index, the stability index, and the steady-state performance index on the control performance of the PID controller, corresponding weight coefficients are assigned;

[0020] The response speed index, the stability index, and the steady-state performance index are subjected to dimensionality reduction processing, and the interaction characteristics of the response speed index and the stability index are analyzed.

[0021] The steady-state performance index is statistically analyzed using a time-series sliding window to obtain the statistical distribution characteristics of the steady-state performance index.

[0022] By combining the interaction features and the statistical distribution features, a multidimensional dynamic behavior feature set of the PID controller is obtained.

[0023] In a preferred embodiment, the step of optimizing the standard temperature feature vector in the parameter space of the PID controller according to an iterative search strategy in the parameter optimization channel to obtain the optimized parameter set of the PID controller includes:

[0024] In the parameter optimization channel, the population of the IWOA algorithm is initialized to generate an initial population containing the parameter combinations in the PID controller;

[0025] The fitness index of the parameter combination is evaluated based on the standard temperature feature vector to obtain the optimal parameter combination of the PID controller;

[0026] In the parameter optimization channel, the position vector of the parameter combination is updated according to the enclosing mechanism of the IWOA algorithm and in combination with the standard temperature feature vector;

[0027] By simulating bubble net attack behavior, the position vector is subjected to local fine processing to obtain the optimized parameter set of the PID controller.

[0028] In a preferred embodiment, updating the position vector of the parameter combination in the parameter optimization channel according to the enclosing mechanism of the IWOA algorithm and in combination with the standard temperature feature vector includes:

[0029] Update the position vector of the parameter combination, wherein the formula for calculating the updated position vector is:

[0030] ;

[0031] in, This represents the updated position vector of the parameter combination. This represents the position vector of the optimal parameter combination. This represents the current position vector of the parameter combination. This represents the standard temperature feature vector. This represents the weighting coefficients of the standard temperature eigenvector. and Represents the coefficient vector;

[0032] The formula for calculating the coefficient vector is as follows:

[0033] ;

[0034] ;

[0035] in, This represents the preset iteration convergence factor. and This represents a random vector.

[0036] In a preferred embodiment, the step of reconfiguring the optimized parameter set into the PID controller through the instruction generation channel, and using the configured PID controller to encode the real-time temperature data sequence to generate preliminary control instructions for the PID controller includes:

[0037] A parameter dynamic loading mechanism is established in the instruction generation channel to write the proportional, integral, and derivative parameters from the optimized parameter set into the parameter storage area of ​​the PID controller.

[0038] The real-time temperature data sequence is standardized by the data format conversion unit of the PID controller to obtain a unified format input data stream for the PID controller.

[0039] Using the dual-channel data processing architecture of the PID controller, the proportional term parameter, the integral term parameter, and the derivative term parameter are quantized and derived to obtain the proportional control component, integral control component, and derivative control component of the PID controller.

[0040] The proportional control component, integral control component, and derivative control component are encoded and synthesized to generate the initial control command of the PID controller.

[0041] In a preferred embodiment, the step of controlling the abrupt change in the amplitude of the initial control command to generate a stable control command for the PID controller, and smoothing the stable control command to obtain the final control command for the PID controller, includes:

[0042] Analyze the historical control data of the PID controller to determine the range of variation of the initial control command;

[0043] The threshold range of the variation range is set according to the system operating status of the PID controller;

[0044] The system tracks the changing trend of the initial control command in real time and limits the amplitude of the initial control command that exceeds the threshold range to the range of change, thereby generating a stable control command for the PID controller.

[0045] In a preferred embodiment, the step of controlling the abrupt change in the amplitude of the initial control command to generate a stable control command for the PID controller, and smoothing the stable control command to obtain the final control command for the PID controller, includes:

[0046] The stable control command is decomposed into control command segments of the PID controller, and local trend fitting is performed on the control command segments to obtain the connection relationship between the control command segments.

[0047] Based on the aforementioned connection relationship, the fitting weight of the control command segment is dynamically adjusted to ensure a natural transition of the control command segment, thereby obtaining the fitted control command of the PID controller.

[0048] Eliminating the inter-segment fluctuations in the fitted control command yields the final control command of the PID controller.

[0049] In a preferred embodiment, the step of triggering the re-optimization condition of the parameter optimization channel when the expected threshold is not reached after executing the final control command, and re-adjusting the PID controller to obtain a PID controller that reaches the expected threshold, includes:

[0050] After executing the final control command, continuously receive the control effect of the final control command and extract the key performance indicators of the control effect;

[0051] The key performance indicators are evaluated in multiple dimensions, and when the key performance indicators are detected to deviate from the expected threshold, a re-optimization trigger signal for the PID controller is automatically generated.

[0052] The re-optimization trigger signal is received through the parameter optimization channel to activate a new round of parameter optimization process and generate the re-optimization parameter set of the PID controller;

[0053] The re-optimized parameter set is reconfigured into the instruction generation channel to adaptively re-regulate the PID controller.

[0054] To address the aforementioned problems, this invention also provides a temperature control system based on IWOA-optimized fuzzy PID, the system comprising:

[0055] The data processing channel construction module is used to generate channels and parameter optimization channels according to the instructions of the PID controller, and to construct the dual-channel data processing architecture of the temperature control system in the PID controller.

[0056] The temperature data processing module is used to extract key indicators reflecting the current dynamic behavior of the PID controller from the real-time temperature data sequence collected by the PID controller, and eliminate the differences in the dimensions and orders of magnitude of the key indicators to obtain the standard temperature feature vector of the PID controller.

[0057] The parameter optimization module is used to optimize the standard temperature feature vector in the parameter space of the PID controller according to the iterative search strategy in the parameter optimization channel to obtain the optimized parameter set of the PID controller.

[0058] The control command generation module is used to reconfigure the optimized parameter set into the PID controller through the command generation channel, and use the configured PID controller to encode the real-time temperature data sequence to generate the initial control command of the PID controller.

[0059] The control command processing module is used to control the amplitude abrupt changes of the initial control command, generate the stable control command of the PID controller, and smooth the stable control command to obtain the final control command of the PID controller.

[0060] The controller retuning module is used to trigger the re-optimization conditions of the parameter optimization channel when the expected threshold is not reached after the execution of the final control command, and to re-tune the PID controller to obtain a PID controller that reaches the expected threshold.

[0061] Compared with the prior art, the present invention has the following beneficial effects:

[0062] 1. This invention constructs a dual-channel data processing architecture to accurately extract key dynamic indicators from real-time temperature data and complete standardization and feature fusion. By combining the encirclement mechanism of the IWOA algorithm with the simulation of bubble net attack behavior, it achieves efficient optimization and fine adjustment of the parameter space, significantly improving the accuracy and response efficiency of PID controller parameter configuration and enhancing the dynamic adaptability of temperature control.

[0063] 2. This invention ensures the stability and continuity of control commands by limiting and smoothing abrupt changes in the amplitude of initial control commands, thereby reducing fluctuations during temperature regulation. The trigger-based re-optimization mechanism can monitor the control effect in real time and promptly initiate a new round of parameter optimization and adaptive regulation, ensuring that temperature control continuously approaches and reaches the expected threshold, thus improving the operational reliability and control quality of the entire control system. Attached Figure Description

[0064] Figure 1 This is a flowchart illustrating a temperature control method based on IWOA-optimized fuzzy PID control according to an embodiment of the present invention.

[0065] Figure 2 A functional block diagram of a temperature control system based on IWOA-optimized fuzzy PID provided in an embodiment of the present invention;

[0066] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0067] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0068] This application provides a temperature control method based on IWOA-optimized fuzzy PID control. The execution entity of this IWOA-optimized fuzzy PID control method includes, but is not limited to, at least one of the following electronic devices that can be configured to execute the method provided in this application: a server, a terminal, etc. In other words, the IWOA-optimized fuzzy PID control method can be executed by software or hardware installed on a terminal device or a server device. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster. The server can be an independent server or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms.

[0069] Reference Figure 1 The diagram shown is a flowchart illustrating a temperature control method based on IWOA-optimized fuzzy PID control according to an embodiment of the present invention. In this embodiment, the temperature control method based on IWOA-optimized fuzzy PID control includes:

[0070] S1. Generate a channel and a parameter optimization channel according to the instructions of the PID controller to construct a dual-channel data processing architecture for the temperature control system in the PID controller;

[0071] The core task of the instruction generation channel is to receive the optimized parameter set output by the parameter optimization channel and convert it into control instructions that can drive the PID controller. This channel needs to be pre-set with a parameter receiving interface, an instruction encoding module, and an instruction output interface. The parameter receiving interface is specifically used to connect to the parameter output end of the parameter optimization channel to ensure that the optimized parameter set can be transmitted completely and without delay. The instruction encoding module is responsible for converting the optimized proportional, integral, and derivative parameters into an instruction format that the PID controller can recognize. The instruction output interface is directly connected to the instruction input end of the PID controller to ensure that the generated control instructions can be delivered accurately.

[0072] The parameter optimization channel uses the real-time temperature data sequence acquired by the PID controller as the processing object. It is equipped with a data acquisition interface, a feature processing module, and a parameter optimization module. The data acquisition interface establishes a stable connection with the temperature data output terminal of the PID controller, captures temperature change data in real time, and forms a continuous real-time temperature data sequence. The feature processing module extracts key indicators and standardizes the real-time temperature data sequence to obtain a standard temperature feature vector that reflects the dynamic behavior of the PID controller. The parameter optimization module selects the optimal parameter combination in the parameter space of the PID controller based on the standard temperature feature vector, forms an optimized parameter set, and transmits it to the instruction generation channel through a preset parameter output interface.

[0073] The instruction generation channel and parameter optimization channel are integrated in a logical sequence of "data acquisition - parameter optimization - instruction generation - controller driving". The hard connection method between the parameter output interface of the parameter optimization channel and the parameter receiving interface of the instruction generation channel is clearly defined to ensure seamless transmission of the optimized parameter set. At the same time, a status feedback mechanism between channels is established. The instruction generation channel feeds back the execution status of the control instructions to the parameter optimization channel in real time. The parameter optimization channel determines whether to start the re-optimization process based on the feedback information. The entire architecture forms a closed-loop data processing flow, ensuring that the temperature control system can dynamically adjust the control strategy according to real-time temperature data.

[0074] The beneficial effects are that this dual-channel data processing architecture, by clearly defining the functional boundaries and collaborative logic of the instruction generation channel and the parameter optimization channel, achieves efficient separation and close cooperation between parameter optimization and instruction generation. This ensures that the optimized parameters can be accurately converted into control instructions, and that the parameter optimization process can be dynamically adjusted based on real-time temperature data. This effectively improves the response speed and stability of the PID controller's temperature control, and provides a structured processing framework for the smooth implementation of subsequent steps, ensuring the orderliness and reliability of the temperature control process.

[0075] S2. Extract key indicators reflecting the current dynamic behavior of the PID controller from the real-time temperature data sequence collected by the PID controller, and eliminate the differences in dimensions and orders of magnitude in the key indicators to obtain the standard temperature feature vector of the PID controller.

[0076] In this embodiment of the invention, the step of extracting key indicators reflecting the current dynamic behavior of the PID controller from the real-time temperature data sequence collected from the PID controller, and eliminating differences in units and orders of magnitude among the key indicators to obtain the standard temperature feature vector of the PID controller, includes:

[0077] The real-time temperature data sequence is divided into multi-scale time windows, and the response speed index, stability index and steady-state performance index of the PID controller are extracted in different time windows respectively.

[0078] The response speed index, the stability index, and the steady-state performance index are fused to obtain a multi-dimensional dynamic behavior feature set of the PID controller;

[0079] Analyze the data distribution of the feature indicators in the multidimensional dynamic behavior feature set to obtain the numerical range and distribution characteristics of the feature indicators;

[0080] Based on the numerical range and the distribution characteristics, the feature index is mapped to a unified interval, and the data distribution pattern of the feature index is adjusted to obtain the standard feature index of the PID controller.

[0081] Correlation analysis is performed on the standard feature indicators, the key feature dimensions of the standard feature indicators are retained, and the key feature dimensions are quantified to obtain the standard temperature feature vector of the PID controller.

[0082] The step of fusing the response speed index, the stability index, and the steady-state performance index to obtain a multi-dimensional dynamic behavior feature set of the PID controller includes:

[0083] According to the degree of influence of the response speed index, the stability index, and the steady-state performance index on the control performance of the PID controller, corresponding weight coefficients are assigned;

[0084] The response speed index, the stability index, and the steady-state performance index are subjected to dimensionality reduction processing, and the interaction characteristics of the response speed index and the stability index are analyzed.

[0085] The steady-state performance index is statistically analyzed using a time-series sliding window to obtain the statistical distribution characteristics of the steady-state performance index.

[0086] By combining the interaction features and the statistical distribution features, a multidimensional dynamic behavior feature set of the PID controller is obtained.

[0087] When dividing real-time temperature data sequences into multi-scale time windows, three fixed-duration time windows are first determined: 1 second, 5 seconds, and 10 seconds. The continuously acquired temperature data is then divided into multiple independent data segments according to these three window durations. Within the 1-second time window, the time interval from the initial deviation of the temperature from the set temperature ±0.5℃ to the start of its return to the set temperature is calculated; this time interval is the response speed index of the PID controller. Within the 5-second time window, the difference between the maximum and minimum values ​​of all temperature data within the window is calculated; this difference is the stability index of the PID controller. Within the 10-second time window, a stable segment with fluctuations of less than 0.1℃ for three consecutive data points within the window is first selected, and then the average difference between the temperature value within the stable segment and the set temperature value is calculated; this average difference is the steady-state performance index of the PID controller.

[0088] When assigning weight coefficients based on the impact of response speed, stability, and steady-state performance on the control performance of the PID controller, the historical operating data of the PID controller for the past 1000 complete temperature control cycles is first retrieved. For each control cycle, the percentage change in the total time for the temperature to reach the set value and remain stable when the response speed index changes by 1%, the percentage change in the total time for the stability index changes by 1%, and the percentage change in the total time for the steady-state performance index changes by 1% are calculated. The average value of these three percentage changes is then calculated. The average percentage change corresponding to the response speed index is the largest, so a weight coefficient of 0.4 is assigned to it; the average percentage change corresponding to the stability index is the second largest, so a weight coefficient of 0.3 is assigned to it; and the average percentage change corresponding to the steady-state performance index is the smallest, so a weight coefficient of 0.3 is assigned to it. This completes the determination of the weight coefficients for each index.

[0089] When performing dimensionality reduction on response speed, stability, and steady-state performance indicators, principal component analysis (PCA) is employed. The values ​​of the three indicators over 1000 historical control cycles are organized into a 3x1000 column data matrix. The covariance matrix of this matrix is ​​calculated, and the eigenvalues ​​and eigenvectors of the covariance matrix are solved. The two eigenvectors with a cumulative contribution rate of over 95% are selected as principal components. Linear combination operations are then performed between the three original indicator data and these two principal components to obtain two dimensionality-reduced principal components. When analyzing the interaction characteristics of the response speed and stability indicators, the product of the first and second principal component values ​​in each control cycle is calculated. Simultaneously, the number of times the first principal component value exceeds the upper limit of the historical normal range and the second principal component value also exceeds the corresponding upper limit of the historical normal range in each control cycle is counted. These two results are used together as the interaction characteristics of the response speed and stability indicators.

[0090] When performing time-series sliding window statistics on steady-state performance indicators, the duration of the sliding window is set to 5 seconds and the sliding step size is 1 second. The steady-state performance indicator data arranged in chronological order within each control cycle are segmented according to the window duration and sliding step size. Within each sliding window, the average, variance, and maximum value of all steady-state performance indicator values ​​within the window are calculated. The average, variance, and maximum value corresponding to all sliding windows are summarized and organized to obtain the statistical distribution characteristics of the steady-state performance indicators.

[0091] When combining interactive features and statistical distribution features, the interactive features of the previously obtained response speed and stability indicators are integrated with the statistical distribution features of the steady-state performance indicators into the same dataset. This dataset contains five feature items: principal component product value, number of out-of-range occurrences, steady-state average deviation, steady-state deviation variance, and steady-state deviation maximum value. This dataset is the multidimensional dynamic behavior feature set of the PID controller.

[0092] When analyzing the data distribution of feature indicators in a multidimensional dynamic behavior feature set, for each feature item in the set, all its values ​​in 1000 historical control cycles are extracted. The minimum and maximum values ​​of each feature item are found to determine the numerical range of that feature item. At the same time, the numerical range of each feature item is divided into 10 intervals on average, and the frequency of the value in each interval is counted. The data distribution characteristics of the feature item are judged based on the frequency distribution, thus obtaining the numerical range and distribution characteristics of each feature indicator.

[0093] Based on the numerical range and distribution characteristics of the characteristic indicators, when mapping the characteristic indicators to a unified interval, a linear normalization method is used, setting the unified interval to [0,1]. For each value of each characteristic item, the value is converted to the [0,1] interval according to the calculation method of "(current value - minimum value of the characteristic item) ÷ (maximum value of the characteristic item - minimum value of the characteristic item)". When adjusting the data distribution of the characteristic indicators, if the data distribution of a certain characteristic item is obviously right-skewed (such as steady-state deviation variance), the normalized value of the characteristic item is transformed by the natural logarithm to make the distribution of the transformed value closer to the normal distribution. The value of each characteristic item after mapping and distribution adjustment is the standard characteristic indicator of the PID controller.

[0094] When performing correlation analysis on standard characteristic indicators, the Pearson correlation coefficient between any two standard characteristic indicators is calculated. The calculation process is as follows: first, calculate the average value of all values ​​of the two indicators, then calculate the deviation of each indicator value from its own average value, multiply the corresponding deviations and sum them, then divide by the square root of the product of the squares of the deviations of the two indicators to obtain the correlation coefficient. If the absolute value of the correlation coefficient between the two standard characteristic indicators is greater than 0.8, then refer to the average change ratio during the previous weight allocation, retain the indicators that have a greater impact on control performance, and remove the indicators with a smaller impact. Finally, retain the three key characteristic dimensions: principal component product value, steady-state average deviation, and steady-state maximum deviation value. When quantifying the key characteristic dimensions, the value of each retained key characteristic dimension is accurate to two decimal places. The quantified values ​​of these three dimensions are combined in sequence into a three-dimensional vector, which is the standard temperature characteristic vector of the PID controller.

[0095] The beneficial effects are as follows: by dividing the time window into multiple scales, response speed, stability, and steady-state performance indicators can be comprehensively extracted from different time dimensions, avoiding the one-sidedness of indicators caused by single-window extraction; during feature fusion, the impact of key indicators on control performance is highlighted by weight allocation, and principal component analysis is used to reduce data redundancy and mine the interaction features between indicators. Time-series sliding window statistics can accurately capture the dynamic distribution law of steady-state performance, ensuring that the multi-dimensional dynamic behavior feature set contains complete and effective information; subsequently, through data distribution analysis, unified interval mapping, and distribution shape adjustment, the differences in dimensions and orders of magnitude are completely eliminated, and the standard temperature feature vector obtained by removing redundant dimensions and quantifying by correlation analysis can accurately reflect the current dynamic behavior of the PID controller, providing a high-quality data foundation for parameter optimization channels, improving the accuracy and efficiency of parameter optimization, and thus ensuring the precision and stability of temperature control.

[0096] S3. In the parameter optimization channel, the standard temperature feature vector is optimized in the parameter space of the PID controller according to the iterative search strategy to obtain the optimized parameter set of the PID controller.

[0097] In this embodiment of the invention, the step of optimizing the standard temperature feature vector in the parameter space of the PID controller according to an iterative search strategy in the parameter optimization channel to obtain the optimized parameter set of the PID controller includes:

[0098] In the parameter optimization channel, the population of the IWOA algorithm is initialized to generate an initial population containing the parameter combinations in the PID controller;

[0099] The fitness index of the parameter combination is evaluated based on the standard temperature feature vector to obtain the optimal parameter combination of the PID controller;

[0100] In the parameter optimization channel, the position vector of the parameter combination is updated according to the enclosing mechanism of the IWOA algorithm and in combination with the standard temperature feature vector;

[0101] By simulating bubble net attack behavior, the position vector is subjected to local fine processing to obtain the optimized parameter set of the PID controller.

[0102] The step of updating the position vector of the parameter combination in the parameter optimization channel according to the encirclement mechanism of the IWOA algorithm and in combination with the standard temperature feature vector includes:

[0103] Update the position vector of the parameter combination, wherein the formula for calculating the updated position vector is:

[0104] ;

[0105] in, This represents the updated position vector of the parameter combination. This represents the position vector of the optimal parameter combination. This represents the current position vector of the parameter combination. This represents the standard temperature feature vector. This represents the weighting coefficients of the standard temperature eigenvector. and Represents the coefficient vector;

[0106] The formula for calculating the coefficient vector is as follows:

[0107] ;

[0108] ;

[0109] in, This represents the preset iteration convergence factor. and This represents a random vector.

[0110] When initializing the population for the IWOA algorithm in the parameter optimization channel, the core control parameters of the PID controller are first determined to be the proportional term parameter P, the integral term parameter I, and the derivative term parameter D. The population size is set to 50 individuals. At the same time, based on the hardware operating limits of the PID controller and historical stable control data, the reasonable value ranges of each parameter are determined: the value range of the proportional term parameter P is 0.1-5.0, the value range of the integral term parameter I is 0.01-2.0, and the value range of the derivative term parameter D is 0.001-1.0. Then, a uniform random number generation method is used to randomly generate corresponding P, I, and D parameter values ​​for each individual in the population within the value range of each parameter. These 50 complete combinations of P, I, and D parameters are organized into a structured data list, which is the initial population containing the PID controller parameter combinations.

[0111] When evaluating the fitness index of parameter combinations based on the standard temperature eigenvector, the fitness index is first defined as the sum of squares of the temperature control deviations of the PID controller. For each P, I, D parameter combination in the initial population, it is temporarily written into the parameter storage area of ​​the PID controller. The PID controller is controlled to run one standard control cycle based on the currently collected real-time temperature data sequence. During the operation, real-time temperature data is collected every 10 seconds, the deviation between the collected temperature and the target temperature is calculated and squared, and all the squared deviation values ​​in the entire cycle are summed to obtain the fitness value of the parameter combination. After calculating the fitness values ​​of all 50 parameter combinations, the size of all fitness values ​​is compared, and the P, I, D parameter combination with the smallest fitness value is selected. This parameter combination is the optimal parameter combination of the PID controller.

[0112] In the parameter optimization channel, when updating the position vector of the parameter combination based on the encirclement mechanism of the IWOA algorithm and the standard temperature feature vector, the P, I, and D parameter values ​​corresponding to each individual in the population are first clarified. The core of the encirclement mechanism is to make the position vector of all non-optimal individuals in the population move closer to the position vector of the optimal parameter combination. During the approach process, the approach magnitude is adjusted in conjunction with the standard temperature feature vector: if the steady-state average deviation in the standard temperature feature vector is greater than 0.5℃, the difference between the non-optimal individual parameter and the optimal parameter is multiplied by an adjustment coefficient of 0.8; if the steady-state average deviation in the standard temperature feature vector is less than or equal to 0.5℃, the difference is multiplied by an adjustment coefficient of 0.3. For example, if the P parameter of a non-optimal individual is 1.6 and the P parameter of the optimal parameter is 2.2, when the deviation is large, the new P value is 1.6 + (2.2 - 1.6) × 0.8 = 2.08; when the deviation is small, the new P value is 1.6 + (2.2 - 1.6) × 0.3 = 1.78. The I and D parameters are adjusted according to the same logic to obtain the updated position vector of each non-optimal individual.

[0113] When performing local fine-tuning on position vectors to simulate bubble net attack behavior, for each updated position vector, a local search range is set around the current value of each parameter. Within this local range, three equally spaced fine-tuning values ​​are generated for each parameter. The fine-tuning values ​​of each parameter are cross-combined to form nine sets of local fine-tuning parameter combinations. The fitness value is calculated for each of these nine sets of fine-tuning combinations. The parameter combination with the smallest fitness value in each set of local fine-tuning combinations is selected as the result of the fine-tuning of the position vector. After completing the local fine-tuning of all position vectors, the fitness values ​​of all fine-tuned parameter combinations are compared again. The P, I, D parameter combination with the smallest fitness value is selected. This parameter combination is the optimized parameter set of the PID controller.

[0114] The position vector of the optimal parameter combination comes from the result obtained after evaluating the fitness index of the parameter combination based on the standard temperature feature vector; the position vector of the current parameter combination comes from the initial population generated by the IWOA algorithm population in the parameter optimization channel, which is the current state before the iterative update; the standard temperature feature vector comes from the real-time temperature data sequence collected by the PID controller, which is obtained after extracting the response speed index, stability index, and steady-state performance index, and then processing it through feature fusion mapping to a unified interval and correlation analysis; the weight coefficient is a preset value used to adjust the degree of influence of the standard temperature feature vector on the position vector update; the iterative convergence factor is a preset value used to control the convergence speed of the IWOA algorithm iteration process; the two random vectors are randomly generated during the parameter optimization process to increase the randomness and diversity of parameter search; the coefficient vector A is obtained by subtracting the iterative convergence factor from the product of twice the iterative convergence factor and the first random vector; the coefficient vector C is obtained by twice the second random vector.

[0115] This calculation process is based on the encirclement mechanism of the IWOA algorithm, and updates the position vector of the parameter combination by combining the standard temperature feature vector. During the calculation, the position vector of the optimal parameter combination and the position vector of the current parameter combination are obtained first, the difference between the two is calculated, and then multiplied by the coefficient vector A. Next, it is multiplied by the reciprocal of the exponential term composed of the standard temperature feature vector and its weight coefficients, and finally the updated position vector is obtained. Its core function is to use the position of the optimal parameter combination to guide the position adjustment of the current parameter combination. At the same time, through the influence of the standard temperature feature vector and the adjustment of the exponential term, the position update of the parameter combination is made to better fit the dynamic requirements of temperature control, realize the optimization search of the PID controller parameter combination, and make the parameter combination closer to the optimal state, thus providing a foundation for the subsequent generation of optimized parameter sets.

[0116] As the iteration process progresses, the iteration convergence factor gradually changes, leading to a change in the numerical range of the coefficient vector A. When the absolute value of the coefficient vector A is less than 1, the position vector of the parameter combination will move closer to the position vector of the optimal parameter combination, the search range will gradually shrink, and it will converge towards the direction of the optimal solution. The larger the magnitude of the standard temperature feature vector, the larger the corresponding exponent value, and the smaller its reciprocal, the stronger the suppression effect on the position vector update, and the smaller the update amplitude of the position vector will be, avoiding over-adjustment. The randomness of the random vector will cause some fluctuations in the position update of the parameter combination, but as the number of iterations increases, the change in the iteration convergence factor will weaken this fluctuation, allowing the position vector update to gradually stabilize. Finally, the position vector of the parameter combination will converge to the vicinity of the position vector of the optimal parameter combination, forming a stable set of optimized parameters.

[0117] The beneficial effects include: clearly defining the parameter range and size during population initialization to ensure that the initial parameter combination covers a reasonable search space and provides sufficient diversity for optimization; fitness evaluation based on the sum of squared temperature deviations combined with complete control cycle data can accurately reflect the actual control effect of parameters and ensure the accuracy of optimal parameter combination selection; the encirclement mechanism, combined with the standard temperature feature vector, dynamically adjusts the proximity amplitude to achieve directional optimization, balancing convergence speed and search rationality; and the local fine processing of simulated bubble net attacks deeply mines better parameters near the optimal position, further improving parameter accuracy. The final optimized parameter set can accurately adapt to the current dynamic behavior of the PID controller, providing a high-quality basis for subsequent parameter configuration and effectively improving the stability and accuracy of temperature control.

[0118] S4. Through the instruction generation channel, the optimized parameter set is reconfigured into the PID controller, and the configured PID controller is used to encode the real-time temperature data sequence to generate the initial control instruction of the PID controller.

[0119] In this embodiment of the invention, the step of reconfiguring the optimized parameter set into the PID controller through the instruction generation channel, and using the configured PID controller to encode the real-time temperature data sequence to generate preliminary control instructions for the PID controller includes:

[0120] A parameter dynamic loading mechanism is established in the instruction generation channel to write the proportional, integral, and derivative parameters from the optimized parameter set into the parameter storage area of ​​the PID controller.

[0121] The real-time temperature data sequence is standardized by the data format conversion unit of the PID controller to obtain a unified format input data stream for the PID controller.

[0122] Using the dual-channel data processing architecture of the PID controller, the proportional term parameter, the integral term parameter, and the derivative term parameter are quantized and derived to obtain the proportional control component, integral control component, and derivative control component of the PID controller.

[0123] The proportional control component, integral control component, and derivative control component are encoded and synthesized to generate the initial control command of the PID controller.

[0124] The method of controlling the abrupt changes in the amplitude of the initial control command generates a stable control command for the PID controller, and smooths the stable control command to obtain the final control command for the PID controller, including:

[0125] Analyze the historical control data of the PID controller to determine the range of variation of the initial control command;

[0126] The threshold range of the variation range is set according to the system operating status of the PID controller;

[0127] The system tracks the changing trend of the initial control command in real time and limits the amplitude of the initial control command that exceeds the threshold range to the range of change, thereby generating a stable control command for the PID controller.

[0128] When establishing a dynamic parameter loading mechanism in the instruction generation channel, the allocation table of the parameter storage area is first read through the hardware debugging interface of the PID controller to determine the independent storage addresses corresponding to the proportional, integral, and derivative parameters. A parameter writing program is developed in the control submodule of the instruction generation channel. This program has built-in parameter validity verification logic, comparing the proportional, integral, and derivative parameters in the optimized parameter set with the allowed extreme value range of the PID controller. If the range is exceeded, a signal to reacquire the optimized parameters is triggered. After successful verification, the program writes the parameter values ​​byte-by-byte to the corresponding storage addresses via the I2C communication protocol in the order of "proportional parameter → integral parameter → derivative parameter". After writing is complete, the parameter values ​​at each address are read back and compared with the written values. Once consistency is confirmed, the parameter configuration in the PID controller's parameter storage area is completed, ensuring accurate loading of the optimized parameters.

[0129] When standardizing the real-time temperature data sequence through the data format conversion unit of the PID controller, the unified format is first set to 32-bit floating-point numbers. The integer part represents the integer in degrees Celsius, and the decimal part retains 2 digits of precision, with a value range of -50.00℃ to 150.00℃. The data format conversion unit first performs anomaly screening on each data point in the real-time temperature data sequence, removing invalid data that is below -60.00℃ or above 160.00℃. Invalid data is replaced with the value of the previous valid data point. Then, following the process of "original data × 100, rounded, converted to 32-bit floating-point number, and then divided by 100", all valid data points are uniformly converted to 32-bit floating-point number format and arranged in chronological order of acquisition time to form a unified format input data stream that can be directly parsed by the PID controller.

[0130] When using a dual-channel data processing architecture of a PID controller to quantize and derive optimized parameters, the instruction generation channel obtains the optimized proportional, integral, and derivative parameters from the parameter optimization channel. Simultaneously, it extracts the deviation between the current temperature value and the preset target temperature value from the unified format input data stream. The deviation is the current temperature value minus the target temperature value. In the quantization derivation module, the proportional control component is calculated using the formula "proportional control component = deviation × optimized proportional parameter," the integral control component is calculated using the formula "integral control component = cumulative deviation of the last 10 times × optimized integral parameter," and the derivative control component is calculated using the formula "derivative control component = (current deviation - previous deviation) / 1 second × optimized derivative parameter." These calculations yield the proportional, integral, and derivative control components of the PID controller.

[0131] When encoding and synthesizing the proportional control component, integral control component, and derivative control component, the encoding rules are first established. Each control component is represented by a 16-bit binary number, with a value range corresponding to the control signal of 0-10V, where 0 corresponds to 0V and 65535 corresponds to 10V. The structure is "proportional component (1-16 bits) + integral component (17-32 bits) + derivative component (33-48 bits) + 8-bit CRC check code (49-56 bits)". The encoding module first converts the three control components into their corresponding 16-bit binary numbers, then concatenates the three binary numbers sequentially. The first 48 bits of data are checked using CRC-8 to generate an 8-bit check code. After concatenation, a 56-bit binary instruction code is formed, which is the initial control instruction of the PID controller.

[0132] When analyzing historical control data of a PID controller to determine the range of variation of the initial control command, the initial control command data of the past 60 days is retrieved from the historical database of the PID controller. This includes the control signal voltage amplitude corresponding to each command. The data statistics module is used to process these amplitude data to calculate the maximum and minimum amplitude values ​​among all the data. The interval formed by the maximum and minimum amplitude values ​​is determined as the range of variation of the initial control command. At the same time, the normal fluctuation frequency of the amplitude within this range is recorded, with fluctuations not exceeding 15 times per minute, to ensure that the range of variation can cover all historical normal command amplitude values.

[0133] When setting the threshold range for the variation range based on the system operating status of the PID controller, the current operating parameters of the PID controller are first collected through the system status monitoring module, including the load rate of the heating / cooling module, ambient temperature, and power supply current. The operating status is divided into low load, medium load, and high load according to the load rate. Low load is defined as a load rate <20%, medium load as a load rate of 20%–60%, and high load as a load rate >60%. Under low load conditions, the threshold range is set to ±12% of the variation range. The lower threshold = lower limit of the variation range + (upper limit of the variation range - lower limit of the variation range) × 12%, and the upper threshold = upper limit of the variation range - (Upper limit of variation range - Lower limit of variation range) × 12%; Under medium load conditions, the threshold range is ±9% of the variation range, the lower threshold = lower limit of variation range + (upper limit of variation range - lower limit of variation range) × 9%, the upper threshold = upper limit of variation range - (upper limit of variation range - lower limit of variation range) × 9%; Under high load conditions, the threshold range is ±6% of the variation range, the lower threshold = lower limit of variation range + (upper limit of variation range - lower limit of variation range) × 6%, the upper threshold = upper limit of variation range - (upper limit of variation range - lower limit of variation range) × 6%, thus obtaining the threshold range of the variation range for the corresponding operating state.

[0134] When tracking the changing trend of the initial control command in real time and generating a stable control command, the command monitoring module collects the voltage amplitude of the control signal corresponding to the initial control command every 0.5 seconds, records the collection time point to form an amplitude change curve to track the trend, and compares each collected amplitude with the threshold range of the current operating state. If the amplitude is lower than the lower threshold, the amplitude is adjusted to the lower threshold; if the amplitude is higher than the upper threshold, it is adjusted to the upper threshold; if the amplitude is within the threshold range, it remains unchanged. The adjusted amplitude is re-encoded into a 56-bit binary command code according to the encoding rules of the initial control command. This command code is the stable control command of the PID controller.

[0135] The beneficial effects are as follows: the dynamic parameter loading mechanism ensures that optimized parameters are accurately written to the PID controller through address confirmation, validity verification, and readback verification, avoiding control failures caused by parameter errors; data format standardization eliminates format differences and abnormal interference in real-time temperature data, providing a unified input for quantization derivation; quantization derivation based on a dual-channel architecture accurately calculates three control components by combining deviation, deviation accumulation, and deviation change rate, ensuring that the components closely match actual temperature regulation requirements; the coded and synthesized preliminary control commands include check bits, improving transmission accuracy; by determining the range of change through historical data and setting thresholds based on operating status, the amplitude of the preliminary command is effectively limited, and the generated stable control commands avoid over-adjustment of the controller, providing a stable foundation for subsequent smoothing processing, ultimately enhancing the continuity and reliability of PID controller temperature control and improving temperature control accuracy.

[0136] S5. Control the abrupt changes in the amplitude of the initial control command, generate a stable control command for the PID controller, and smooth the stable control command to obtain the final control command for the PID controller.

[0137] In this embodiment of the invention, the step of controlling the abrupt change in the amplitude of the initial control command to generate a stable control command for the PID controller, and smoothing the stable control command to obtain the final control command for the PID controller, includes:

[0138] The stable control command is decomposed into control command segments of the PID controller, and local trend fitting is performed on the control command segments to obtain the connection relationship between the control command segments.

[0139] Based on the aforementioned connection relationship, the fitting weight of the control command segment is dynamically adjusted to ensure a natural transition of the control command segment, thereby obtaining the fitted control command of the PID controller.

[0140] Eliminating the inter-segment fluctuations in the fitted control command yields the final control command of the PID controller.

[0141] When decomposing the stable control command into control command segments for the PID controller, the stable control command is first determined to be a 56-bit binary command code. Each command code corresponds to the control signal voltage amplitude at a given moment. Continuous stable control commands are collected in a time sequence, and every 10 consecutive stable control commands form an independent control command segment. Each command segment contains 10 control signal voltage amplitude data points. When performing local trend fitting on each control command segment, a linear regression method is used, with time unit as the horizontal axis and control signal voltage amplitude as the vertical axis. The linear regression equation for the 10 data points within each command segment is calculated. By comparing the slopes of the linear regression equations of two adjacent control command segments, the connection relationship is determined. If the slope difference between adjacent segments is less than 0.2V / s, it is determined to be a smooth connection relationship; if the slope difference is greater than or equal to 0.2V / s, it is determined to be a steep connection relationship. Thus, the connection relationship between the control command segments of the PID controller is obtained.

[0142] When dynamically adjusting the fitting weights of control command segments based on their connection relationships, initial fitting weights are first set. The fitting weight for data points at non-connection points within each control command segment is 1.0, and the initial weight for data points at the connection points of adjacent control command segments is 1.0. If the connection relationship is smooth, the weight of the data points at the connection point is kept at 1.0. If the connection relationship is steep, the fitting weight of the data points at the connection point is adjusted to 1.5, and the fitting weight of one data point before and after the connection point is adjusted to 1.2. Based on the adjusted fitting weights, a weighted linear trend fitting is performed again on each control command segment to obtain a smooth fitting curve for each command segment. The smooth fitting curves of all command segments are then spliced ​​together in chronological order to obtain the fitted control command of the PID controller.

[0143] To eliminate inter-segment fluctuations in the fitted control command, the connection points of the fitted curves of two adjacent control command segments are first extracted. This involves finding the amplitude point corresponding to the last time unit of the previous fitted curve and the amplitude point corresponding to the first time unit of the next fitted curve, and calculating the amplitude difference between the two connection points. If the amplitude difference is greater than 0.1V, five transition amplitude points with equal time intervals are inserted between the two connection points. The amplitude of the transition amplitude points is calculated according to the rule "amplitude of the previous connection point + (amplitude of the next connection point - amplitude of the previous connection point) × (transition point number / 6)" to achieve a smooth amplitude transition. If the amplitude difference is less than or equal to 0.1V, the two connection points are directly linearly connected. After completing the connection point processing, a three-point moving average is applied to the entire fitted control command sequence. The amplitude of each time unit is equal to the average of the amplitude of that time unit and the amplitude of the time units before and after it. After processing, the final control command of the PID controller is obtained.

[0144] The beneficial effects are as follows: by decomposing stable control commands according to time series and combining linear regression for local trend fitting, the variation patterns of each command segment and the connection status of adjacent segments can be accurately identified, providing a clear basis for subsequent smoothing processing; by dynamically adjusting the fitting weights based on the connection relationship, the data influence at steep connection points can be specifically strengthened, avoiding abrupt transitions between segments caused by fitting deviations and ensuring natural transitions between command segments; by judging the amplitude difference and inserting transition points and using moving average processing, the inter-segment fluctuations of the fitted control commands can be effectively eliminated, giving the final control commands continuous and smooth variation characteristics, avoiding sudden temperature rises and falls when the PID controller executes commands, significantly improving the stability and accuracy of temperature control, and ensuring that the controlled object is in a stable temperature environment.

[0145] S6. When the expected threshold is not reached after the final control command is executed, the re-optimization condition of the parameter optimization channel is triggered to re-adjust the PID controller so as to obtain a PID controller that reaches the expected threshold.

[0146] In this embodiment of the invention, when the expected threshold is not reached after executing the final control command, triggering the re-optimization condition of the parameter optimization channel to re-adjust the PID controller to obtain a PID controller that reaches the expected threshold includes:

[0147] After executing the final control command, continuously receive the control effect of the final control command and extract the key performance indicators of the control effect;

[0148] The key performance indicators are evaluated in multiple dimensions, and when the key performance indicators are detected to deviate from the expected threshold, a re-optimization trigger signal for the PID controller is automatically generated.

[0149] The re-optimization trigger signal is received through the parameter optimization channel to activate a new round of parameter optimization process and generate the re-optimization parameter set of the PID controller;

[0150] The re-optimized parameter set is reconfigured into the instruction generation channel to adaptively re-regulate the PID controller.

[0151] When continuously receiving the control effect of the final control command after its execution and extracting the key performance indicators of the control effect, a high-precision temperature sensor is first deployed at the temperature monitoring point of the controlled object. The sensor's sampling frequency is 1 time per second. The actual temperature data of the controlled object is collected in real time by the sensor. This actual temperature data is the control effect data of the final control command. Three key performance indicators are extracted from the control effect data: temperature stability, temperature deviation, and response time. Temperature stability is obtained by calculating the difference between the maximum and minimum values ​​of the actual temperature data within 5 consecutive minutes. Temperature deviation is obtained by calculating the average of the absolute values ​​of the difference between the actual temperature and the preset target temperature within 5 consecutive minutes. Response time is obtained by recording the time interval from the start of the final control command execution to the actual temperature first entering the preset target temperature ±0.5℃ range and remaining within 1 minute. This completes the extraction of the key performance indicators of the control effect.

[0152] When performing multi-dimensional performance evaluation of key performance indicators and automatically generating a re-optimization trigger signal for the PID controller when a key performance indicator deviates from the expected threshold, the expected thresholds for each key performance indicator are first preset. The expected threshold for temperature stability is set to ≤0.3℃, the expected threshold for temperature deviation is set to ≤±0.2℃, and the expected threshold for response time is set to ≤30 seconds. A parallel evaluation method is adopted, and the extracted temperature stability, temperature deviation, and response time are compared with their corresponding expected thresholds simultaneously. If any key performance indicator exceeds its corresponding expected threshold, such as temperature stability of 0.4℃, temperature deviation of ±0.3℃, or response time of 35 seconds, it is determined that the key performance indicator deviates from the expected threshold. At this time, a high-level 3.3V re-optimization trigger signal is generated in the signal generation module of the parameter optimization channel, and the signal duration is 1 second to ensure that the parameter optimization channel can reliably capture the re-optimization trigger signal.

[0153] When the parameter optimization channel receives the re-optimization trigger signal and activates a new round of parameter optimization to generate the re-optimized parameter set for the PID controller, the signal receiving module of the parameter optimization channel monitors the re-optimization trigger signal in real time. When a high-level re-optimization trigger signal lasting for 1 second is detected, a new round of parameter optimization is automatically activated. The new round of parameter optimization still uses the IWOA algorithm. When initializing the population of the IWOA algorithm, the value range of each parameter is narrowed to ±30% of the current parameter value based on the parameter combination currently used by the PID controller. For example, when the current proportional parameter is 2.0, the new value range is 1.4-2.6, and the population size remains unchanged at 50 individuals. A standard temperature feature vector is generated based on the latest collected actual temperature data sequence. The fitness index of each parameter combination in the population is evaluated based on the standard temperature feature vector. Then, the position vector of the parameter combination is updated according to the encirclement mechanism of the IWOA algorithm. At the same time, the position vector is locally refined by simulating bubble net attack behavior. Finally, the parameter combination with the smallest fitness value is selected, and this parameter combination is the re-optimized parameter set of the PID controller.

[0154] When the re-optimized parameter set is reconfigured into the instruction generation channel and the PID controller is adaptively readjusted, the instruction generation channel initiates a dynamic parameter loading mechanism. It reads the proportional, integral, and derivative parameters from the re-optimized parameter set and writes these parameters byte-by-byte into the parameter storage area of ​​the PID controller via the I2C communication protocol. After the parameters are written, the parameter values ​​at each storage address are read back to verify consistency with the written re-optimized parameter values. After the verification is successful, the instruction generation channel standardizes the real-time acquired temperature data sequence based on the re-optimized parameters, and then derives the proportional, integral, and derivative control components through quantization. These three control components are encoded and synthesized to obtain a new preliminary control instruction. Then, the amplitude abrupt change of the preliminary control instruction is controlled to generate a stable control instruction. The stable control instruction is smoothed to obtain a new final control instruction. The new final control instruction is sent to the PID controller for execution. At the same time, new control effects are continuously received and key performance indicators are extracted. The above process of multi-dimensional performance evaluation, parameter re-optimization, and instruction regeneration is repeated until the key performance indicators reach the expected threshold, thus completing the adaptive readjustment of the PID controller.

[0155] The beneficial effects include: continuous acquisition of actual temperature data through high-precision temperature sensors to obtain control effects, which can reflect the execution status of the final control command in real time and accurately; the extracted multi-dimensional key performance indicators can comprehensively cover the stability, accuracy and response speed of temperature control, avoiding the omission of control problems caused by single-indicator evaluation; multi-dimensional performance evaluation combined with clear expected thresholds can accurately determine whether the control effect meets the standard, ensuring the timeliness and accuracy of the re-optimization trigger signal generation, and avoiding the aggravation of temperature deviation due to delayed triggering; the new round of parameter optimization narrows the search range based on the current parameters, which can reduce the optimization time of the IWOA algorithm, improve efficiency, and make the re-optimized parameter set more in line with the current control scenario, ensuring parameter adaptability; adaptive re-tuning, through the closed-loop operation of dynamic parameter loading and command generation, ensures that the re-optimized parameters take effect quickly and continuously optimize the control effect, ultimately enabling the PID controller to stably reach the expected threshold, significantly improving the robustness and accuracy of temperature control, and meeting the needs of the controlled object for a stable temperature environment.

[0156] like Figure 2 The diagram shown is a functional block diagram of a temperature control system based on IWOA optimized fuzzy PID provided in an embodiment of the present invention.

[0157] The temperature control system 100 based on IWOA optimized fuzzy PID, as described in this invention, can be installed in an electronic device. Depending on the functions implemented, the temperature control system 100 based on IWOA optimized fuzzy PID may include a data processing channel construction module 101, a temperature data processing module 102, a parameter optimization module 103, a control command generation module 104, a control command processing module 105, and a controller re-regulation module 106. The module described in this invention can also be called a unit, referring to a series of computer program segments that can be executed by the processor of an electronic device and perform a fixed function, stored in the memory of the electronic device.

[0158] In this embodiment, the functions of each module / unit are as follows:

[0159] The data processing channel construction module 101 is used to generate channels and parameter optimization channels according to the instructions of the PID controller, and to construct a dual-channel data processing architecture for the temperature control system in the PID controller.

[0160] The temperature data processing module 102 is used to extract key indicators reflecting the current dynamic behavior of the PID controller from the real-time temperature data sequence collected by the PID controller, and eliminate the differences in the dimensions and orders of magnitude of the key indicators to obtain the standard temperature feature vector of the PID controller.

[0161] The parameter optimization module 103 is used to optimize the standard temperature feature vector in the parameter space of the PID controller according to an iterative search strategy in the parameter optimization channel to obtain the optimized parameter set of the PID controller.

[0162] The control command generation module 104 is used to reconfigure the optimized parameter set into the PID controller through the command generation channel, and use the configured PID controller to encode the real-time temperature data sequence to generate the initial control command of the PID controller.

[0163] The control command processing module 105 is used to control the amplitude abrupt change of the preliminary control command, generate the stable control command of the PID controller, and smooth the stable control command to obtain the final control command of the PID controller.

[0164] The controller retuning module 106 is used to trigger the re-optimization conditions of the parameter optimization channel when the expected threshold is not reached after the execution of the final control command, and to re-tune the PID controller to obtain a PID controller that reaches the expected threshold.

[0165] In the several embodiments provided by this invention, it should be understood that the disclosed methods and systems can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and other division methods may be used in actual implementation.

[0166] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0167] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional modules.

[0168] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0169] This application embodiment can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence is the theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0170] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A temperature control method based on IWOA-optimized fuzzy PID, characterized in that, The method includes: S1. Generate a channel and a parameter optimization channel according to the instructions of the PID controller to construct a dual-channel data processing architecture for the temperature control system in the PID controller; S2. Extract key indicators reflecting the current dynamic behavior of the PID controller from the real-time temperature data sequence collected by the PID controller, and eliminate the differences in dimensions and orders of magnitude in the key indicators to obtain the standard temperature feature vector of the PID controller. S3. In the parameter optimization channel, the standard temperature feature vector is optimized in the parameter space of the PID controller according to the iterative search strategy to obtain the optimized parameter set of the PID controller. S4. Through the instruction generation channel, the optimized parameter set is reconfigured into the PID controller, and the configured PID controller is used to encode the real-time temperature data sequence to generate the initial control instruction of the PID controller. S5. Control the abrupt changes in the amplitude of the initial control command, generate a stable control command for the PID controller, and smooth the stable control command to obtain the final control command for the PID controller. S6. When the expected threshold is not reached after the final control command is executed, the re-optimization condition of the parameter optimization channel is triggered to re-adjust the PID controller so as to obtain a PID controller that reaches the expected threshold.

2. The temperature control method based on IWOA optimized fuzzy PID as described in claim 1, characterized in that, The process involves extracting key indicators reflecting the current dynamic behavior of the PID controller from the real-time temperature data sequence acquired from the PID controller, and eliminating differences in dimensions and orders of magnitude among these key indicators to obtain the standard temperature feature vector of the PID controller, including: The real-time temperature data sequence is divided into multi-scale time windows, and the response speed index, stability index and steady-state performance index of the PID controller are extracted in different time windows respectively. The response speed index, the stability index, and the steady-state performance index are fused to obtain a multi-dimensional dynamic behavior feature set of the PID controller; Analyze the data distribution of the feature indicators in the multidimensional dynamic behavior feature set to obtain the numerical range and distribution characteristics of the feature indicators; Based on the numerical range and the distribution characteristics, the feature index is mapped to a unified interval, and the data distribution pattern of the feature index is adjusted to obtain the standard feature index of the PID controller. Correlation analysis is performed on the standard feature indicators, the key feature dimensions of the standard feature indicators are retained, and the key feature dimensions are quantified to obtain the standard temperature feature vector of the PID controller.

3. The temperature control method based on IWOA optimized fuzzy PID as described in claim 2, characterized in that, The step of fusing the response speed index, the stability index, and the steady-state performance index to obtain a multi-dimensional dynamic behavior feature set of the PID controller includes: According to the degree of influence of the response speed index, the stability index, and the steady-state performance index on the control performance of the PID controller, corresponding weight coefficients are assigned; The response speed index, the stability index, and the steady-state performance index are subjected to dimensionality reduction processing, and the interaction characteristics of the response speed index and the stability index are analyzed. The steady-state performance index is statistically analyzed using a time-series sliding window to obtain the statistical distribution characteristics of the steady-state performance index. By combining the interaction features and the statistical distribution features, a multidimensional dynamic behavior feature set of the PID controller is obtained.

4. The temperature control method based on IWOA optimized fuzzy PID as described in claim 1, characterized in that, In the parameter optimization channel, the standard temperature feature vector is optimized in the parameter space of the PID controller according to an iterative search strategy to obtain the optimized parameter set of the PID controller, including: In the parameter optimization channel, the population of the IWOA algorithm is initialized to generate an initial population containing the parameter combinations in the PID controller; The fitness index of the parameter combination is evaluated based on the standard temperature feature vector to obtain the optimal parameter combination of the PID controller; In the parameter optimization channel, the position vector of the parameter combination is updated according to the enclosing mechanism of the IWOA algorithm and in combination with the standard temperature feature vector; By simulating bubble net attack behavior, the position vector is subjected to local fine processing to obtain the optimized parameter set of the PID controller.

5. The temperature control method based on IWOA optimized fuzzy PID as described in claim 4, characterized in that, The step of updating the position vector of the parameter combination in the parameter optimization channel according to the encirclement mechanism of the IWOA algorithm and in combination with the standard temperature feature vector includes: Update the position vector of the parameter combination, wherein the formula for calculating the updated position vector is: ； in, This represents the updated position vector of the parameter combination. This represents the position vector of the optimal parameter combination. This represents the current position vector of the parameter combination. This represents the standard temperature feature vector. This represents the weighting coefficients of the standard temperature eigenvector. and Represents the coefficient vector; The formula for calculating the coefficient vector is as follows: ； ； in, This represents the preset iteration convergence factor. and This represents a random vector.

6. The temperature control method based on IWOA optimized fuzzy PID as described in claim 1, characterized in that, The step involves reconfiguring the optimized parameter set into the PID controller via the instruction generation channel, and then using the configured PID controller to encode the real-time temperature data sequence to generate preliminary control instructions for the PID controller, including: A parameter dynamic loading mechanism is established in the instruction generation channel to write the proportional, integral, and derivative parameters from the optimized parameter set into the parameter storage area of ​​the PID controller. The real-time temperature data sequence is standardized by the data format conversion unit of the PID controller to obtain a unified format input data stream for the PID controller. Using the dual-channel data processing architecture of the PID controller, the proportional term parameter, the integral term parameter, and the derivative term parameter are quantized and derived to obtain the proportional control component, integral control component, and derivative control component of the PID controller. The proportional control component, integral control component, and derivative control component are encoded and synthesized to generate the initial control command of the PID controller.

7. The temperature control method based on IWOA optimized fuzzy PID as described in claim 1, characterized in that, The method of controlling the abrupt changes in the amplitude of the initial control command generates a stable control command for the PID controller, and smooths the stable control command to obtain the final control command for the PID controller, including: Analyze the historical control data of the PID controller to determine the range of variation of the initial control command; The threshold range of the variation range is set according to the system operating status of the PID controller; The system tracks the changing trend of the initial control command in real time and limits the amplitude of the initial control command that exceeds the threshold range to the range of change, thereby generating a stable control command for the PID controller.

8. The temperature control method based on IWOA optimized fuzzy PID as described in claim 1, characterized in that, The method of controlling the abrupt changes in the amplitude of the initial control command generates a stable control command for the PID controller, and smooths the stable control command to obtain the final control command for the PID controller, including: The stable control command is decomposed into control command segments of the PID controller, and local trend fitting is performed on the control command segments to obtain the connection relationship between the control command segments. Based on the aforementioned connection relationship, the fitting weight of the control command segment is dynamically adjusted to ensure a natural transition of the control command segment, thereby obtaining the fitted control command of the PID controller. Eliminating the inter-segment fluctuations in the fitted control command yields the final control command of the PID controller.

9. The temperature control method based on IWOA optimized fuzzy PID as described in claim 1, characterized in that, When the expected threshold is not reached after executing the final control command, the re-optimization condition of the parameter optimization channel is triggered to re-adjust the PID controller to obtain a PID controller that reaches the expected threshold, including: After executing the final control command, continuously receive the control effect of the final control command and extract the key performance indicators of the control effect; The key performance indicators are evaluated in multiple dimensions, and when the key performance indicators are detected to deviate from the expected threshold, a re-optimization trigger signal for the PID controller is automatically generated. The re-optimization trigger signal is received through the parameter optimization channel to activate a new round of parameter optimization process and generate the re-optimization parameter set of the PID controller; The re-optimized parameter set is reconfigured into the instruction generation channel to adaptively re-regulate the PID controller.

10. A temperature control system based on IWOA-optimized fuzzy PID, used to implement the temperature control method based on IWOA-optimized fuzzy PID as described in claim 1, the system comprising: The data processing channel construction module is used to generate channels and parameter optimization channels according to the instructions of the PID controller, and to construct the dual-channel data processing architecture of the temperature control system in the PID controller. The temperature data processing module is used to extract key indicators reflecting the current dynamic behavior of the PID controller from the real-time temperature data sequence collected by the PID controller, and eliminate the differences in the dimensions and orders of magnitude of the key indicators to obtain the standard temperature feature vector of the PID controller. The parameter optimization module is used to optimize the standard temperature feature vector in the parameter space of the PID controller according to the iterative search strategy in the parameter optimization channel to obtain the optimized parameter set of the PID controller. The control command generation module is used to reconfigure the optimized parameter set into the PID controller through the command generation channel, and use the configured PID controller to encode the real-time temperature data sequence to generate the initial control command of the PID controller. The control command processing module is used to control the amplitude abrupt changes of the initial control command, generate the stable control command of the PID controller, and smooth the stable control command to obtain the final control command of the PID controller. The controller retuning module is used to trigger the re-optimization conditions of the parameter optimization channel when the expected threshold is not reached after the execution of the final control command, and to re-tune the PID controller to obtain a PID controller that reaches the expected threshold.

Images