High-precision and quick-response temperature adjusting system for industrial thermostat

Abstract

The invention relates to the technical field of industrial automation control, and discloses a high-precision and quick-response industrial thermostat temperature regulation system which comprises a temperature acquisition module, a fluctuation analysis module, a dynamic calibration module, a gradient optimization module, a threshold management module and a regulation execution module. The temperature acquisition module acquires real-time temperature and environment thermal inertia parameters, and sets a temperature adjustment interval; the fluctuation analysis module divides temperature nodes and generates fluctuation feature vectors; the dynamic calibration module extracts a temperature deviation index and establishes an adjustment compensation rule; the gradient optimization module identifies a temperature change grade and calculates a fluctuation density difference value; the threshold value management module deduces an optimal adjustment threshold value and generates a threshold value deviation sequence; and the adjustment execution module is integrated to form a temperature adjustment execution scheme. The system realizes deep analysis and dynamic compensation of temperature fluctuation through cooperation of multiple modules, improves the adjustment precision and the response speed, and is suitable for high-precision temperature control industrial scenes such as chemical engineering and mechanical manufacturing.

Description

Technical Field

[0001] The invention relates to the technical field of industrial automation control, and in particular to a high-precision, fast-response industrial thermostat temperature regulation system. Background Art

[0002] In the industrial production process, temperature is a key process parameter, and its precise control directly affects the equipment operation stability, production efficiency and product quality. As the core component of temperature control, industrial thermostats are widely used in chemical industry, machinery manufacturing, energy and power and other fields. They are responsible for real-time temperature adjustment of target equipment to ensure that the equipment operates within a safe and efficient temperature range. However, the temperature control system of existing industrial thermostats generally has problems such as insufficient adjustment accuracy, slow response speed, and poor environmental adaptability, which makes it difficult to meet the needs of modern industry for high-precision and high-reliability temperature control.

[0003] From the perspective of technical implementation, traditional temperature control systems usually use a fixed threshold PID control strategy, the core defect of which is that they cannot dynamically adapt to changes in the equipment's operating status and environmental thermal inertia. For example, when the equipment load changes suddenly or the ambient temperature fluctuates drastically, the fixed control parameters can easily lead to temperature regulation lags, resulting in large temperature overshoots or regulation blind spots, which in turn cause equipment failures or production accidents. In addition, the traditional system's analysis of temperature fluctuation data remains at a simple statistical level, lacking the ability to deeply analyze and recognize the fluctuation characteristics of temperature nodes, making it difficult to predict temperature change trends in advance and formulate targeted regulation strategies.

[0004] In actual applications, the complexity of industrial scenarios further increases the difficulty of temperature regulation. Different devices have unique thermal characteristics at different operating stages. For example, parameters such as thermal inertia strength and temperature change rate vary significantly. Traditional systems often adopt a "one-size-fits-all" adjustment mode, which cannot achieve accurate matching with the operating status of the equipment. At the same time, when multiple devices are running in coordination, the thermal coupling effect between the devices will lead to uneven distribution of the temperature field. Traditional systems lack the ability to dynamically calibrate and optimize gradients for global temperature fluctuations, which can easily lead to systematic temperature control deviations.

[0005] With the development of industrial automation and intelligence, high-precision and fast-response temperature control has become a key requirement for industry upgrades. For example, in temperature-sensitive fields such as semiconductor manufacturing and precision instrument processing, temperature fluctuations must be controlled within ±0.1°C, and traditional regulation systems can no longer meet such stringent requirements. Therefore, developing a new temperature regulation system that can sense equipment status and environmental changes in real time, deeply analyze temperature fluctuation characteristics, and dynamically optimize regulation strategies has become a technical problem that needs to be solved in the field of industrial thermostat technology.

[0006] In the prior art, although some research attempts to introduce machine learning algorithms to analyze temperature data, most of them remain at the laboratory stage, and face problems such as high algorithm complexity, poor real-time performance, and expensive hardware costs in practical applications. In addition, the existing system lacks an effective linkage mechanism in the threshold management and adjustment execution links, resulting in a large disconnect between the adjustment instruction and the actual temperature deviation, and it is difficult to achieve the optimization of closed-loop control. Summary of the Invention

[0007] The purpose of the present invention is to provide an industrial thermostat temperature regulation system with high precision and fast response to solve the problems mentioned in the above background technology.

[0008] To achieve the above purpose, the present invention provides the following technical solutions: an industrial thermostat temperature regulation system with high precision and fast response, the system includes: A temperature acquisition module, used to obtain the real-time temperature data and environmental thermal inertia parameters of the target device, and set a temperature regulation interval matching the device operation state, and the temperature regulation interval is the temperature fluctuation data to be processed; A fluctuation analysis module, used to divide multiple temperature nodes within the temperature regulation interval, perform pattern recognition on the fluctuation data of each temperature node, and generate a fluctuation feature vector corresponding to the temperature node; A dynamic calibration module, used to extract temperature deviation indicators from the fluctuation feature vector, establish an adjustment compensation rule associated with the temperature node, and obtain the threshold control parameters corresponding to the adjustment compensation rule; A gradient optimization module, used to identify the temperature change level in the threshold control parameters, dynamically compensate the temperature deviation indicators according to the temperature change level, and calculate the fluctuation density difference of each temperature node under different adjustment strategies; A threshold management module, used to derive the optimal adjustment threshold according to the fluctuation density difference, and generate a threshold deviation sequence by comparing the current temperature fluctuation density with the optimal adjustment threshold; An adjustment execution module, used to analyze the threshold deviation sequence, and integrate the threshold deviation sequence into a temperature adjustment execution plan based on the fluctuation density characteristics of the temperature node.

[0009] Preferably, the implementation method of the fluctuation analysis module includes: constructing a device feature library corresponding to the temperature node, and the device feature library includes a temperature parameter vector mapped by real-time temperature data and environmental thermal inertia parameters; Perform similar fluctuation matching on the temperature parameter vector, divide the fluctuation clustering group of the temperature parameter vector according to the matching result; extract the distribution balance point of the temperature data from the fluctuation clustering group, and set the distribution balance point as the temperature node.

[0010] Preferably, dividing the fluctuation clustering group of the temperature parameter vector further includes: Extract the temperature change rate, adjustment interval, and thermal inertia strength parameters based on the temperature control frequency and thermal inertia fluctuation index in the temperature parameter vector, and generate a fluctuation feature label based on the temperature change rate, adjustment interval, and thermal inertia strength parameters; Associate the fluctuation feature label with the temperature parameter vector, and filter the parameter vectors with a similarity higher than the preset fluctuation threshold to form a fluctuation clustering group by calculating the fluctuation similarity between the feature labels.

[0011] Preferably, the implementation method for generating the fluctuation feature vector corresponding to the temperature node includes: For each temperature node, obtain the fluctuation difference parameter of the temperature node within a preset period according to the time sequence distribution of the temperature node in the temperature adjustment interval, and calculate the temperature difference coefficient of the node; When the temperature difference coefficient exceeds the first calibration threshold, mark the node as an unbalanced node and extract its temperature data to form a fluctuation feature vector; when the temperature difference coefficient is lower than the first calibration threshold, mark the node as a stable node, and perform gradient superposition on the temperature data of the adjacent nodes of the node, and reconstruct the superimposed data into a fluctuation feature vector.

[0012] Preferably, the implementation method of the dynamic calibration module includes: Separate the temperature data ratio, abnormal fluctuation ratio, and thermal inertia fluctuation parameters from the fluctuation feature vector, and generate an adjustment compensation rule for the temperature node based on the above parameters; If the number of temperature nodes covered by the current adjustment compensation rule is less than the preset fluctuation threshold, traverse the fluctuation feature vectors of the adjacent temperature nodes, and add the temperature indicators not included in the adjustment compensation rules of the adjacent nodes to the current rule.

[0013] Preferably, the implementation method of the gradient optimization module includes: obtaining the time sequence parameter of the adjustment validity period and the fluctuation parameter of the gradient compensation strength in the temperature change level; Construct a compensation state network associated with the time sequence parameter and the fluctuation parameter, and determine the fluctuation density difference under different adjustment strategies according to the switching probability of various paths in the compensation state network.

[0014] Preferably, constructing the compensation state network further includes: Identify the periodic pattern of the time sequence parameter. If the current periodic pattern completely matches the preset temperature control period, set the time sequence parameter as the starting point of the compensation state network; Calculate the gradient correlation degree between the time sequence parameter and the fluctuation parameter, and generate the intermediate point and the end point of the compensation state network in descending order of the gradient correlation degree; Perform a state reverse verification on the end point. When the gradient correlation degree of the end point is lower than the preset gradient threshold, output it as the final link of the compensation state network.

[0015] Preferably, the implementation method for calculating the fluctuation density difference includes: Statistically calculate the mean value of the timing parameters and the range of the fluctuation parameters of each termination point in the compensation status network, and calculate the global covariance of all node parameters; Subtract the mean value of the timing parameters of a single termination point from the mean value of the timing parameters of the adjacent node, divide the obtained difference by the global covariance to obtain the timing difference coefficient; at the same time, calculate the ratio of the range of the fluctuation parameters to the global covariance, and weight and sum the ratio and the timing difference coefficient as the fluctuation density difference of this node.

[0016] Preferably, the implementation method for deriving the optimal adjustment threshold includes: Extract the adjustment mode in the historical data that is closest to the current fluctuation density difference, calculate the Euclidean distance in the timing distribution between the fluctuation density difference in the closest adjustment mode and the current fluctuation density difference as the first threshold reference value; Statistically calculate the difference in the number of peak points between the fluctuation density difference in the closest adjustment mode and the current fluctuation density difference, and use the difference quantity as the second threshold reference value; Based on the linear combination of the first threshold reference value and the second threshold reference value, match the optimal adjustment threshold in the preset adjustment threshold table.

[0017] Preferably, the implementation method of the adjustment execution module includes: dividing the positive fluctuation domain and the reverse fluctuation domain according to the temperature fluctuation trend of each node in the threshold deviation sequence; Extract the convergence frequency of the threshold deviation in the positive fluctuation domain and the diffusion frequency of the threshold deviation in the reverse fluctuation domain, and dynamically reconcile the two according to the adjustment weight of the temperature node to generate the configuration parameters of the temperature adjustment execution plan.

[0018] Compared with the prior art, the beneficial effects of the present invention are: The temperature acquisition module can obtain the temperature data of the target device and the environmental thermal inertia parameters in real time, and set a matching temperature adjustment range according to the device operation state, realizing the accurate positioning and preprocessing of the temperature fluctuation data. This design breaks through the limitation of the fixed threshold of the traditional system, enabling the adjustment system to dynamically adapt to the thermal characteristic requirements of different operation stages of the device, and laying a data foundation for subsequent high-precision adjustment.

[0019] The fluctuation analysis module realizes the in-depth deconstruction of the temperature parameter vector by constructing an equipment feature library, performing similar fluctuation matching, and clustering analysis. This module can not only divide temperature nodes and generate corresponding fluctuation feature vectors, but also generate fluctuation feature labels by extracting parameters such as temperature change rate and adjustment interval, effectively identifying the patterns and rules of temperature fluctuations. This refined analysis ability enables the system to predict the temperature change trend in advance, providing a key basis for dynamic calibration and gradient optimization, and significantly enhancing the system's adaptability to complex temperature fluctuation scenarios.

[0020] The dynamic calibration module extracts the temperature deviation index from the fluctuation feature vector, establishes an adjustment compensation rule associated with the temperature node, and dynamically expands the rule coverage by traversing the data of adjacent nodes. This mechanism ensures that the adjustment strategy can respond to the subtle changes in temperature fluctuations in real time. Especially when the number of temperature nodes is small or the fluctuation characteristics are not obvious, through the supplementary analysis of adjacent node data, it avoids the one-sidedness and lag of the adjustment rule, and improves the comprehensiveness and accuracy of the adjustment compensation.

[0021] The gradient optimization module dynamically calculates the fluctuation density difference under different adjustment strategies by constructing a compensation state network and combining the gradient correlation degree of the time series parameters and the fluctuation parameters. This module can not only identify the temperature change level, but also optimize the adjustment strategy according to the path switching probability of the compensation state network, realizing the dynamic compensation of the temperature deviation index. This calculation method based on the global covariance and the time series difference coefficient enables the system to select the optimal solution among multiple adjustment strategies, significantly enhancing the dynamic response speed and adjustment accuracy of the temperature adjustment.

[0022] The threshold management module generates the optimal adjustment threshold by analyzing the adjustment patterns in the historical data, combining the Euclidean distance and the peak point difference, and generates a deviation sequence by comparing the current fluctuation density with the threshold. This design realizes the dynamic update and adaptive adjustment of the threshold, avoiding the problem of adjustment failure of the traditional fixed threshold in complex scenarios, and enabling the system to always maintain the best adjustment state under different environments and equipment states.

[0023] The adjustment execution module divides the fluctuation domain according to the threshold deviation sequence, extracts the convergence frequency and the diffusion frequency, and dynamically reconciles them to generate configuration parameters. This module converts the abstract deviation data into a specific adjustment execution plan, realizing the precise matching of the adjustment instruction and the temperature fluctuation trend, and ensuring the effectiveness and timeliness of the adjustment action. The closed-loop linkage mechanism between multiple modules forms a complete control loop from data acquisition to adjustment execution, significantly enhancing the overall control performance of the system. Description of the Drawings

[0024] Figure 1 It is the working principle diagram of the high-precision and fast-response industrial thermostat temperature adjustment system described in the present invention; Figure 2 Flow chart for dividing fluctuation clustering groups; Figure 3 Flow chart for generating fluctuation feature vectors; Figure 4 Flow chart for calculating fluctuation density difference. Specific implementation manners

[0025] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0026] Please refer to Figures 1-4 , the high-precision and fast-response industrial thermostat temperature regulation system involved in the present invention, which includes a temperature acquisition module, a fluctuation analysis module, a dynamic calibration module, a gradient optimization module, a threshold management module, and an adjustment execution module. The specific steps are as follows: Temperature acquisition module: Real-time collect the temperature data of the target device through a temperature sensor array, and at the same time obtain the environmental thermal inertia parameters (such as environmental temperature, humidity, air flow velocity, etc.) through the environmental parameter monitoring unit. Based on the device operation state parameters (such as power output, load level), set a temperature regulation interval that matches the device operation state through a preset state-temperature mapping algorithm, and the temperature fluctuation data within this interval is used as the object for subsequent processing.

[0027] Fluctuation analysis module: Divide multiple temperature nodes within the temperature regulation interval. Perform pattern recognition on the fluctuation data of each node to generate corresponding fluctuation feature vectors. The specific implementation is: Extract time-domain and frequency-domain features through algorithms such as Fourier transform and wavelet analysis, and construct a feature vector containing parameters such as mean value, variance, and peak frequency.

[0028] Dynamic calibration module: Extract temperature deviation indicators (such as absolute deviation, relative deviation, cumulative deviation) from the fluctuation feature vectors, and establish an adjustment compensation rule associated with the temperature nodes. The rule is based on the PID control principle and combines a fuzzy logic algorithm to dynamically adjust the compensation coefficient according to the deviation indicators. At the same time, obtain the threshold control parameters (such as proportional coefficient, integral time, differential coefficient) corresponding to the rule.

[0029] Gradient optimization module: Identify the temperature change levels (such as low-speed change, medium-speed change, high-speed change) in the recognition threshold control parameters, and dynamically compensate the temperature deviation index according to the levels. Through the Particle Swarm Optimization (PSO) algorithm or genetic algorithm, calculate the difference in fluctuation density (i.e., the difference in the number of temperature fluctuations per unit time) of each temperature node under different adjustment strategies (such as segmented control, adaptive control).

[0030] Threshold management module: Derive the optimal adjustment threshold through regression analysis or neural network model based on the difference in fluctuation density. Compare the current temperature fluctuation density with the optimal threshold to generate a threshold deviation sequence containing the deviation values of each node.

[0031] Adjustment execution module: Analyze the threshold deviation sequence. Based on the fluctuation density characteristics of each node, convert the deviation sequence into specific adjustment actions through actuators (such as electric control valves, cooling fans), such as adjusting the coolant flow rate, starting and stopping the cooling device, etc., to form a closed-loop control.

[0032] The present invention will be further described below in conjunction with Embodiments 1 to 5: Embodiment 1: The specific implementation methods of the fluctuation analysis module include links such as the construction of the device feature library, the processing of temperature parameter vectors, the division of fluctuation clustering groups, and the setting of temperature nodes. Each link realizes the structured analysis of temperature fluctuations through technical means such as data mapping, pattern matching, and feature extraction, as follows: Construct a device feature library corresponding to the temperature nodes. The device feature library is a comprehensive database established based on the historical operation data and real-time monitoring data of the target device. Its core function is to store multi-dimensional data information related to the temperature nodes. At the data acquisition level, the temperature data of the target device is collected in real time through temperature sensors (such as thermocouples, thermal resistors) distributed at key parts of the device. The acquisition frequency can be set to once per second or higher according to the operating characteristics of the device to ensure the timeliness and continuity of the data. At the same time, the environmental thermal inertia parameters, including environmental temperature, humidity, air flow velocity, heat conduction coefficient, etc., are synchronously obtained through the environmental parameter monitoring unit (such as temperature and humidity sensors, anemometers). These parameters are used to characterize the degree of influence of the environment on the temperature change of the device.

[0033] In the data mapping stage, the real-time temperature data is associated and mapped with the environmental thermal inertia parameters to generate a temperature parameter vector. Each temperature parameter vector corresponds to a specific time point and contains dimensions such as timestamp, real-time temperature value, environmental thermal inertia parameter value, etc. For example, the temperature parameter vector at a certain moment can be expressed as [May 29, 2025 10:00:00, 45°C, environmental temperature 25°C, humidity 60%, air flow velocity 0.5 m / s, heat transfer coefficient 0.3 W / (m·K)]. Among them, the timestamp is used to identify the data collection moment, the real-time temperature value reflects the current temperature state of the device, and the environmental thermal inertia parameter is used for subsequent fluctuation analysis.

[0034] Perform similar fluctuation matching on the temperature parameter vectors. The purpose of similar fluctuation matching is to classify the temperature parameter vectors with similar fluctuation patterns for subsequent clustering analysis. In this process, the Dynamic Time Warping (DTW) algorithm is used to calculate the similarity between different temperature parameter vectors. The DTW algorithm is a method for measuring the similarity of two time series data. Its core idea is to align the time points of the two sequences by bending the time axis to find the optimal matching path. Specifically, for two temperature parameter vector sequences, the algorithm constructs a distance matrix, and each element in the matrix represents the distance between the corresponding dimensions of the two vectors (such as Euclidean distance). Then, through dynamic programming, the shortest path from the starting point to the ending point of the matrix is found, and the length of this path is the similarity metric value between the two sequences. The smaller the similarity metric value, the more similar the fluctuation patterns of the two temperature parameter vectors.

[0035] According to the results of similar fluctuation matching, the temperature parameter vectors are divided into different fluctuation clustering groups. The fluctuation clustering group is a set of vectors with similar temperature fluctuation patterns. The vectors within each clustering group have high consistency in terms of temperature change trend, fluctuation amplitude, frequency, etc. When dividing the clustering groups, a preset fluctuation threshold is set (such as the similarity metric value is less than 0.5), and the temperature parameter vectors with similarity higher than this threshold are grouped into the same clustering group. For example, if the similarity metric value between vector A and vector B is 0.3, which is less than the preset threshold of 0.5, then they are grouped into the same clustering group; if the similarity metric value between vector C and vector A is 0.6, which is greater than the preset threshold, then it is grouped into a different clustering group.

[0036] After completing the division of the fluctuation clustering groups, it is necessary to extract the distribution equilibrium points of the temperature data from each clustering group and set them as temperature nodes. The distribution equilibrium point is a characteristic value representing the central tendency of the temperature data within the clustering group, which can be determined by calculating the mean, median, or mode of the temperature data within the clustering group. For example, for the temperature data sequence [42°C, 43°C, 45°C, 44°C, 43°C] within a certain clustering group, its mean is 43.4°C and the median is 43°C. One of these values can be used as the distribution equilibrium point of this clustering group. After setting the distribution equilibrium point as the temperature node, these nodes will serve as the key reference points for subsequent temperature fluctuation analysis and adjustment, used to divide the temperature adjustment interval and generate the corresponding fluctuation feature vectors.

[0037] Furthermore, when dividing the fluctuation clustering groups of the temperature parameter vector, two parameters, namely the temperature control frequency and the thermal inertia fluctuation index, also need to be considered. The temperature control frequency refers to the number of times the temperature control system performs adjustment actions per unit time, reflecting the response frequency of the system to temperature fluctuations; the thermal inertia fluctuation index is an index measuring the change rate of the environmental thermal inertia parameter, and its calculation formula is (the current moment thermal inertia parameter value - the previous moment thermal inertia parameter value) / time interval, used to characterize the change speed of the environmental thermal inertia.

[0038] Based on the temperature control frequency, the thermal inertia fluctuation index, and other parameters in the temperature parameter vector, the temperature change rate, the adjustment interval, and the thermal inertia intensity parameter are extracted. The temperature change rate is the change amount of the equipment temperature per unit time, and its calculation formula is (the current moment temperature value - the previous moment temperature value) / time interval, used to describe the speed of temperature fluctuation; the adjustment interval is the time difference between two adjacent temperature adjustment actions, reflecting the time interval characteristics of the system adjustment; the thermal inertia intensity parameter is the absolute value of the environmental thermal inertia parameter, used to represent the degree of obstruction of the environment to the change of the equipment temperature.

[0039] According to the above three parameters (temperature change rate, adjustment interval, thermal inertia intensity parameter), fluctuation feature labels are generated. The fluctuation feature label is an abstract description of the temperature fluctuation pattern. For example, when the temperature change rate is low (such as less than 1°C / min), the adjustment interval is long (such as greater than 10 min), and the thermal inertia intensity parameter is large (such as greater than 0.5 W / (m·K)), a fluctuation feature label of "low frequency, low speed, high inertia" can be generated; when the temperature change rate is high (such as greater than 5°C / min), the adjustment interval is short (such as less than 2 min), and the thermal inertia intensity parameter is small, a fluctuation feature label of "high frequency, high speed, low inertia" can be generated.

[0040] After generating the fluctuation feature labels, associate them with the temperature parameter vectors so that each vector carries the corresponding fluctuation feature label. Then, screen the parameter vectors with a fluctuation similarity higher than the preset fluctuation threshold by calculating the fluctuation similarity between the feature labels to form a fluctuation clustering group. The calculation of the fluctuation similarity can use the cosine similarity algorithm, which measures the similarity between two feature label vectors by calculating the cosine value of the included angle between them. For example, for the two feature labels "low frequency, low speed, high inertia" and "low frequency, medium speed, high inertia", after converting them into vector form and calculating the cosine value, if the cosine value is greater than the preset threshold (such as 0.7), it is considered that they have a high similarity, and the corresponding temperature parameter vectors can be classified into the same fluctuation clustering group.

[0041] Through the above steps, the fluctuation analysis module can effectively perform structured analysis on the temperature fluctuation data within the temperature adjustment range, divide the complex temperature fluctuation patterns into multiple temperature nodes with clear characteristics, and provide accurate input data for subsequent modules such as dynamic calibration and gradient optimization, thereby realizing the high-precision and fast response characteristics of the industrial thermostat temperature adjustment system. The entire process is data-driven, combined with pattern recognition and feature extraction technologies, ensuring the scientificity and rationality of the temperature node division and laying a foundation for the improvement of the overall performance of the system.

[0042] Embodiment 2: The implementation method of generating the fluctuation feature vector corresponding to the temperature node is carried out around the timing distribution of the temperature node within the temperature adjustment range. Through links such as collecting the fluctuation difference parameters, calculating the temperature difference coefficient, determining the node state, and generating the feature vector, the feature extraction and reconstruction of temperature data under different fluctuation states are realized. The specific process is as follows: For each temperature node, based on its timing distribution within the temperature adjustment range (that is, the node sequence arranged in chronological order), obtain the fluctuation difference parameters of the node within the preset period. The preset period can be set according to the frequency characteristics of the equipment temperature fluctuation. For example, for equipment with relatively slow temperature changes, the preset period can be set to 10 minutes; for equipment with relatively fast temperature changes, the preset period can be shortened to 1 minute. Within the preset period, continuously sample the equipment temperature through a temperature sensor, and the sampling interval matches the accuracy of the sensor and the control requirements of the equipment, usually once per second or once every few seconds.

[0043] The fluctuation difference parameter is the absolute difference between the temperature values at adjacent times within the preset period, which is used to reflect the fluctuation amplitude of the temperature in a short time. For example, the sampling data of a certain temperature node within the preset period is [40°C, 42°C, 41°C, 43°C, 42°C], then the fluctuation difference parameters at adjacent times are 2°C, 1°C, 2°C, 1°C in sequence. By statistically analyzing the distribution of these difference parameters, the fluctuation characteristics of the temperature node can be further analyzed.

[0044] On the basis of obtaining the fluctuation difference parameter, calculate the temperature difference coefficient of the node. The temperature difference coefficient is a key indicator to measure the stability of temperature fluctuations, and its calculation method is the ratio of the standard deviation to the mean value of the fluctuation difference parameter. The standard deviation reflects the degree of dispersion of the fluctuation difference parameter, and the mean value reflects the average level of the fluctuation difference. The ratio of the two can effectively characterize the relative stability of temperature fluctuations. For example, if the mean value of the fluctuation difference parameter of a certain node is 1.5°C and the standard deviation is 0.5°C, then the temperature difference coefficient is 0.5 / 1.5≈0.33; if the mean value of the fluctuation difference parameter of another node is 1.0°C and the standard deviation is 0.8°C, then the temperature difference coefficient is 0.8 / 1.0 = 0.8, and the fluctuation stability of the latter is lower than that of the former.

[0045] According to the comparison result between the temperature difference coefficient and the first calibration threshold, determine the state of the temperature node. The first calibration threshold is a preset critical value for distinguishing the node fluctuation state, and its value can be determined according to the temperature control accuracy requirements of the device and the statistical results of historical data. For example, it can be set to 1.0 or 1.5. When the temperature difference coefficient exceeds the first calibration threshold, it indicates that the temperature fluctuation amplitude of this node is large and the stability is poor. At this time, mark this node as an unbalanced node; when the temperature difference coefficient is lower than the first calibration threshold, it indicates that the temperature fluctuation amplitude of this node is small and the stability is good. At this time, mark this node as a stable node.

[0046] For the temperature nodes marked as unbalanced nodes, directly extract their temperature data to form a fluctuation feature vector. The extracted temperature data is usually the continuous sampling values within a preset period, such as the temperature values of 20 consecutive sampling points. These data contain the temperature fluctuation details of this node in the near future and can intuitively reflect its abnormal fluctuation state. The dimension of the fluctuation feature vector can be determined according to the subsequent analysis requirements. For example, it includes a timestamp, temperature value, fluctuation difference parameter, etc., so that the dynamic calibration module and the gradient optimization module can conduct in-depth analysis on it.

[0047] For the temperature nodes marked as stable nodes, it is necessary to perform gradient superposition processing on the temperature data of their adjacent nodes. Adjacent nodes refer to the previous node and the subsequent node adjacent to this stable node in the chronological distribution of the temperature adjustment interval. Gradient superposition is to perform weighted summation on the temperature data of adjacent nodes according to the preset weights. For example, set the weight of the previous node to 0.3, the weight of the current stable node to 0.4, and the weight of the subsequent node to 0.3, and perform superposition calculation on the temperature data of the three nodes to obtain the superimposed data sequence. This processing method can make full use of the temperature data information around the stable node to make up for the deficiencies of the data characteristics of a single stable node.

[0048] After completing the gradient superposition, it is necessary to reconstruct the superimposed data to generate a fluctuation feature vector. The reconstruction process uses data dimensionality reduction techniques, such as principal component analysis (PCA). By mapping the high-dimensional temperature data to a low-dimensional space, redundant information in the data is removed, and the principal components that can best reflect the temperature fluctuation characteristics are extracted. For example, the superimposed three-dimensional temperature data (previous node, current node, next node) is reduced to a two-dimensional or one-dimensional feature vector through PCA. This vector contains the main fluctuation trends and correlation information of the original data and can effectively characterize the overall fluctuation characteristics of stable nodes and their adjacent nodes.

[0049] During the entire implementation process, the following points need to be noted: First, the setting of the preset period should be flexible and can be dynamically adjusted according to changes in the device operating state. For example, when there is a sudden change in the device load, the preset period is automatically shortened to increase the data acquisition frequency; Second, the determination of the first calibration threshold should be based on a large amount of historical data statistics and device control requirement analysis to avoid inaccurate determination of node states due to improper threshold setting; Third, the weight distribution of gradient superposition should be reasonably designed according to the position and importance of the node in the time series distribution. For example, the weight of the node close to the current stable node can be appropriately increased to highlight its leading role; Fourth, sufficient feature information should be retained during the data dimensionality reduction process to ensure that the reconstructed fluctuation feature vector can accurately reflect the fluctuation pattern of the original data.

[0050] Through the above method, the generated fluctuation feature vector can effectively distinguish the temperature fluctuation characteristics of unbalanced nodes and stable nodes. For unbalanced nodes, the original temperature data is directly used to extract features to ensure the integrity of abnormal fluctuation information; for stable nodes, through the gradient superposition and dimensionality reduction reconstruction of adjacent node data, an effective characterization of the normal fluctuation pattern is achieved. This differential processing method can not only capture abnormal situations in temperature fluctuations but also extract the laws of normal fluctuations, providing accurate input data for the subsequent dynamic calibration module. It helps the industrial thermostat temperature regulation system formulate corresponding adjustment strategies according to different fluctuation characteristics, thereby achieving the temperature control goals of high precision and fast response. The entire process ensures the scientificity and effectiveness of the fluctuation feature vector through data processing and feature engineering techniques, providing strong support for the improvement of the overall performance of the system.

[0051] Example 3: The implementation method of the dynamic calibration module focuses on the parameter analysis of the fluctuation feature vector, the generation of adjustment compensation rules, and the rule extension mechanism. Through data separation, logical judgment, and rule fusion technologies, the dynamic optimization of the temperature node adjustment strategy is realized. The specific process is as follows: Separate the proportion of temperature data, the proportion of abnormal fluctuations, and the thermal inertia fluctuation parameters from the fluctuation feature vector. The fluctuation feature vector is a multi-dimensional data set output by the temperature acquisition module and the fluctuation analysis module, including dimensions such as temperature values, timestamps, environmental thermal inertia parameters, and fluctuation differences. The proportion of temperature data is the ratio of the sum of the temperature value dimensions in the vector to the sum of all dimensions of the vector, which is used to measure the weight of temperature values in the overall fluctuation characteristics; the proportion of abnormal fluctuations is the ratio of the number of temperature fluctuations exceeding the preset warning threshold in the vector to the total number of fluctuations. The preset warning threshold is determined according to the temperature range of normal operation of the device. For example, if the normal temperature range of the device is [30°C, 50°C], the warning threshold can be set below 30°C or above 50°C; the thermal inertia fluctuation parameter is the change amount of environmental thermal inertia parameters (such as environmental temperature, humidity, air flow velocity) within a preset time period, and the calculation formula is:

[0052] where, represents the thermal inertia fluctuation parameter, represents the value of the environmental thermal inertia parameter at the current moment, represents the value of the environmental thermal inertia parameter time units ago, and is the number of time units of the preset time period (such as represents the previous 10 minutes).

[0053] Based on the separated proportion of temperature data, the proportion of abnormal fluctuations, and the thermal inertia fluctuation parameters, generate adjustment compensation rules associated with temperature nodes. The adjustment compensation rules adopt a condition-action logical structure. For example: when the proportion of temperature data is greater than 60% and the proportion of abnormal fluctuations is greater than 20%, trigger the compensation action of increasing the proportional coefficient by 20%; when the thermal inertia fluctuation parameter is greater than 0.5 and the proportion of temperature data is less than 40%, trigger the compensation action of shortening the integral time by 15%. The generation of these rules is based on the historical data of device operation and control theory, such as the combination of PID control principle and fuzzy logic algorithm. By analyzing the temperature adjustment effects under different parameter combinations, a mapping relationship between parameter thresholds and compensation actions is established.

[0054] After generating the adjustment compensation rule, it is necessary to determine whether the number of temperature nodes covered by the current rule is less than the preset fluctuation threshold. The preset fluctuation threshold is determined according to the total number of temperature nodes that need to be adjusted by the system. For example, when there are 20 temperature nodes in the system, the preset fluctuation threshold can be set to 6 (i.e., 30% of the total number of nodes). If the number of nodes covered by the current rule is less than this threshold, it means that the applicable range of the existing rule is relatively narrow and cannot meet the adjustment requirements of complex temperature scenarios. At this time, it is necessary to traverse the fluctuation feature vectors of adjacent temperature nodes and extract temperature indicators not included in the adjustment compensation rules of adjacent nodes, such as peak temperature, valley temperature, maximum and minimum values of temperature change rate, etc.

[0055] The process of extracting temperature indicators is as follows: for the fluctuation feature vector of each adjacent node, analyze the temporal distribution of its temperature data, determine the peak point and valley point of the temperature curve, the peak temperature is the temperature value at the peak point, and the valley temperature is the temperature value at the valley point; calculate the maximum and minimum values of the temperature change rate, that is, the maximum value of the temperature rise rate and the minimum value of the temperature drop rate per unit time. These indicators can further describe the detailed characteristics of temperature fluctuations. For example, the peak temperature reflects the upper limit of temperature fluctuations, and the maximum value of the temperature change rate reflects the situation of rapid temperature rise.

[0056] After extracting the new temperature indicators, add them to the current adjustment compensation rule through the rule fusion algorithm. The rule fusion algorithm uses the method of decision tree integration, takes the existing rules and the newly extracted indicators as input features, and generates a composite rule containing the new indicators by training multiple decision tree models. For example, the existing rule is "when the proportion of temperature data > 60% and the proportion of abnormal fluctuations > 20%, increase by 20%", and the newly extracted indicator is peak temperature > 55°C, then the fused rule can be extended to "when the proportion of temperature data > 60% and the proportion of abnormal fluctuations > 20% and peak temperature > 55°C, increase by 30%", to cope with higher temperature peaks by increasing the compensation amplitude.

[0057] During the rule extension process, the following points need to be noted: First, the extraction of temperature indicators should be targeted, and indicators related to the scenarios where the current rule coverage is insufficient should be selected preferentially. For example, when the existing rule is insufficient in dealing with high-temperature peak scenarios, focus on extracting peak temperature indicators; second, the training data of the rule fusion algorithm should contain sufficient historical cases to ensure that the generated composite rule has statistical significance; third, the compensation action amplitude of the new rule should be determined through sensitivity analysis to avoid deterioration of the temperature adjustment effect due to excessive or insufficient compensation; fourth, the extended rules need to be subjected to consistency verification to ensure that there are no logical conflicts between different rules, such as not being able to trigger both increase and decrease actions under the same parameter combination.

[0058] Through the above steps, the dynamic calibration module can generate an initial adjustment compensation rule based on the parameter characteristics of the fluctuation feature vector, and dynamically improve the rule through the index extension of adjacent nodes when the rule coverage is insufficient. This mechanism enables the system to adapt to the fluctuation characteristics of different temperature nodes, especially to make precise adjustment responses to complex and changeable temperature scenarios (such as sudden changes in equipment load and drastic changes in environmental thermal inertia). The separation of the temperature data ratio, abnormal fluctuation ratio, and thermal inertia fluctuation parameters provides a multi-dimensional input basis for rule generation; while the rule fusion algorithm and index extension mechanism ensure the comprehensiveness and flexibility of the adjustment strategy, avoiding adjustment lags or deficiencies caused by a single rule. The entire process realizes the dynamic optimization of the adjustment compensation rule through a combination of data-driven and logical reasoning, laying a foundation for the high-precision control of the industrial thermostat temperature regulation system.

[0059] Embodiment 4: The implementation method of the gradient optimization module is based on the temperature change level in the threshold control parameter. By extracting the timing parameter of the adjustment validity period and the fluctuation parameter of the gradient compensation intensity, constructing a compensation state network and calculating the fluctuation density difference, the quantitative evaluation of different adjustment strategies is realized. The specific process is as follows: Obtain the timing parameter and fluctuation parameter in the temperature change level. The temperature change level is divided into low-speed change (such as ≤1°C / min), medium-speed change (1 - 5°C / min), and high-speed change (≥5°C / min) according to the rate of equipment temperature fluctuation, and each level corresponds to a different adjustment strategy. The timing parameter of the adjustment validity period refers to the duration from the execution of the adjustment action to the occurrence of an obvious temperature change. For example, after a certain cooling fan is started, the effective time for the equipment temperature to start to drop is 2 minutes; the fluctuation parameter of the gradient compensation intensity refers to the change range of the compensation coefficient in the adjustment strategy. For example, in the high-speed change level, the proportional coefficient may be adjusted from 0.5 to 0.8, and its fluctuation parameter is 0.3.

[0060] When constructing a compensation state network associated with the above parameters, it is first necessary to identify the periodic pattern of the timing parameter. The periodic pattern refers to the regular characteristics of the timing parameter changing with time. For example, the equipment shows a daily periodic pattern of temperature rise from 9 am to 11 am every day due to increased load, or an hourly periodic pattern of temperature fluctuation. By analyzing the timing distribution of historical data, if the current periodic pattern completely matches the preset temperature control period (such as the typical operation period of the equipment) (such as both are a fluctuation cycle of 2 hours), then set this timing parameter as the starting point of the compensation state network. For example, take the moment when the temperature starts to rise as the starting point of the network.

[0061] Calculate the gradient correlation degree of the timing parameter and the fluctuation parameter. The gradient correlation degree is used to measure the correlation between two parameters. For example, in a scenario with a longer adjustment validity period, whether the gradient compensation intensity needs to be adjusted accordingly. When calculating, statistical methods are used to analyze the change trends of the two. For example, when the adjustment validity period is extended, if the gradient compensation intensity also shows an increasing trend, it is considered that the two are positively correlated; if it shows a decreasing trend, it is considered negatively correlated. Generate the intermediate points and termination points of the compensation state network in descending order of correlation degree. The intermediate points represent the key states in the parameter change process (such as the compensation intensity change node when the validity period is extended from 2 minutes to 5 minutes), and the termination points represent the final states of the parameter change (such as the compensation intensity value when the validity period is stable at 10 minutes).

[0062] When performing state reverse verification on the termination point, it is necessary to trace back from the termination point to the starting point to check whether the gradient correlation degree of the entire path is logical. For example, if the compensation intensity corresponding to a certain termination point shows a contradiction with the change trend of the validity period during the backtracking process (such as the validity period shortens but the compensation intensity continues to increase), and the correlation degree is lower than the preset gradient threshold (such as lower than 0.4), then it is determined that this path is invalid and needs to be regenerated; if the correlation degree meets the requirements, it is output as the final link of the compensation state network. The final link clearly shows the parameter change path from the starting point to the termination point, such as "starting point (validity period 2 minutes, compensation intensity 0.2) → intermediate point (validity period 5 minutes, compensation intensity 0.5) → termination point (validity period 10 minutes, compensation intensity 0.8)", providing a path basis for the subsequent calculation of the fluctuation density difference.

[0063] When calculating the fluctuation density difference, first calculate the mean value of the timing parameters and the range of the fluctuation parameters for each termination point in the compensation state network. The mean value of the timing parameters is the average of all the timing parameters in the path corresponding to this termination point. For example, if a termination point path includes validity periods of 2 minutes, 5 minutes, and 10 minutes, its mean value is (2 + 5 + 10) / 3 ≈ 5.67 minutes; the range of the fluctuation parameters is the difference between the maximum and minimum values of the fluctuation parameters in this path. For example, if the compensation intensity changes from 0.2 to 0.8, the range is 0.6. At the same time, calculate the global covariance of all node parameters (including the timing parameters and fluctuation parameters of the starting point, intermediate points, and termination points). The global covariance is used to measure the overall correlation between all parameters and reflects the degree of cooperation of parameter changes under different adjustment strategies.

[0064] Then, the mean of the timing parameters of a single termination point is subtracted from the mean of the timing parameters of the adjacent node to obtain the timing difference. For example, the mean of termination point A is 5.67 minutes, the mean of adjacent termination point B is 8 minutes, and the difference is -2.33 minutes. This difference is divided by the global covariance to obtain the timing difference coefficient, which is used to standardize the timing difference and eliminate the influence of dimensions. At the same time, the ratio of the range of the fluctuation parameter to the global covariance is calculated to obtain the fluctuation difference coefficient, which reflects the significance of the change in the fluctuation parameter relative to the overall correlation. The timing difference coefficient and the fluctuation difference coefficient are weighted and summed according to a preset weight (such as 60% for timing and 40% for fluctuation) to obtain the fluctuation density difference of this node. For example, if the timing difference coefficient is -0.5, the fluctuation difference coefficient is 0.8, and the weights are 0.6 and 0.4 respectively, then the fluctuation density difference is (-0.5×0.6)+(0.8×0.4)=-0.3 + 0.32 = 0.02. This value represents the degree of difference in the temperature fluctuation density under this adjustment strategy and the adjacent strategy.

[0065] Taking an industrial equipment as an example, its temperature change level is medium-speed change (3°C / min), and the preset temperature control period is 1 hour. During the construction of the compensation state network, it is recognized that the current timing parameters show an obvious hourly cycle pattern, and the starting point is set at the 10th minute of the temperature rising stage (the start time of the validity period). By analyzing historical data, three main paths are extracted: Path 1: Starting point (validity period 10 minutes, compensation intensity 0.3) → Intermediate point (validity period 25 minutes, compensation intensity 0.5) → Termination point (validity period 40 minutes, compensation intensity 0.7), the timing mean is 25 minutes, and the fluctuation range is 0.4; Path 2: Starting point (validity period 10 minutes, compensation intensity 0.3) → Intermediate point (validity period 30 minutes, compensation intensity 0.6) → Termination point (validity period 50 minutes, compensation intensity 0.9), the timing mean is 30 minutes, and the fluctuation range is 0.6; Path 3: Starting point (validity period 10 minutes, compensation intensity 0.3) → Intermediate point (validity period 15 minutes, compensation intensity 0.4) → Termination point (validity period 20 minutes, compensation intensity 0.5), the timing mean is 15 minutes, and the fluctuation range is 0.2.

[0066] When calculating the global covariance, the time series parameters (10, 25, 40, 30, 50, 15, 20) and volatility parameters (0.3, 0.5, 0.7, 0.6, 0.9, 0.4, 0.5) of all nodes are integrated, and the covariance value of the two is obtained as 12.5. Taking the end point of Path 1 as an example, the difference in the time series mean between it and the end point of Path 2 is 25 - 30 = -5 minutes. Dividing by the covariance of 12.5 gives a time series difference coefficient of -0.4; the ratio of the volatility range of 0.4 to the covariance is 0.4 / 12.5 = 0.032. Calculated according to the weights of 0.6 and 0.4, the volatility density difference is (-0.4 × 0.6) + (0.032 × 0.4) = -0.24 + 0.0128 = -0.2272, indicating that the volatility density of Path 1 is lower than that of Path 2, which may correspond to a more stable adjustment effect.

[0067] In practical applications, the construction of the compensation status network needs to be dynamically updated. For example, when the device replacement batch or environmental parameters change, the node parameters and link relationships are adjusted through newly collected real-time data. The calculation of the gradient correlation degree needs to combine the physical characteristics of the device. For example, in a device with high heat conduction efficiency, the correlation degree between the adjustment validity period and the compensation intensity may be higher, and the weight of the time series parameters needs to be increased. The comparison of the volatility density differences can help the system automatically screen the optimal adjustment strategy. For example, in the above example, if the goal is to reduce the volatility density, the system can preferentially select Path 3 with a smaller volatility density difference because its time series mean is shorter and the volatility amplitude is smaller, which may be more suitable for the requirements of rapid response.

[0068] Through parameter correlation analysis and network modeling, the entire process transforms the abstract adjustment strategy into a quantifiable volatility density difference, providing a scientific decision-making basis for the threshold management module. The dynamic construction and verification of the compensation status network ensure the timeliness and accuracy of the model, while the difference calculation method of weighted summation comprehensively considers the dual effects of time series and volatility, enabling the system to dynamically optimize the adjustment strategy according to real-time data, and ultimately achieving high-precision control of the industrial thermostat under different temperature change levels.

[0069] Example 5: The implementation method of deriving the optimal adjustment threshold and generating the temperature adjustment execution plan realizes the precise mapping from the volatility density difference to the specific adjustment action based on historical data matching, parameter difference analysis, and dynamic reconciliation algorithm. The specific process is as follows: To derive the optimal adjustment threshold, it is necessary to extract from historical data the adjustment mode that is closest to the current difference in fluctuation density. The historical data is stored in the system database. Each adjustment mode includes the difference in fluctuation density, the corresponding adjustment threshold, and the temporal distribution characteristics of temperature fluctuations. For example, a certain historical adjustment mode is recorded as: the difference in fluctuation density is 0.15, the adjustment threshold is 45°C, and the temperature fluctuations show a trend of being higher in the morning and lower in the afternoon in terms of time sequence. Calculate the similarity between the current difference in fluctuation density and the historical mode using the Euclidean distance. The Euclidean distance formula is the square root of the sum of the squares of the differences in each dimension. In this scenario, there is only one dimension (the difference in fluctuation density), so the distance is the absolute value difference. Suppose the current difference is 0.18, and there are three close values 0.17, 0.19, and 0.21 in the historical mode. Among them, the distance of 0.19 is the smallest (0.01), so this mode is selected as the closest reference.

[0070] When calculating the Euclidean distance (i.e., the first threshold reference value) in the temporal distribution between the closest adjustment mode and the current difference in fluctuation density, it is necessary to align the temperature fluctuation data of the two according to time points. For example, the current data records the sequence of the difference in fluctuation density per hour within 24 hours [0.15, 0.18, 0.20,...], and the corresponding sequence of the historical mode is [0.16, 0.19, 0.21,...]. Calculate the square root of the sum of the squares of the differences at each time point to obtain the first threshold reference value. At the same time, count the difference in the number of peak points between the two (i.e., the second threshold reference value). The peak point is defined as a local maximum point (e.g., the difference at a certain moment is greater than the values at the previous and subsequent moments). If the current data has 5 peak points and the historical mode has 4, then the difference in quantity is 1.

[0071] Based on the linear combination of the first threshold reference value and the second threshold reference value (such as weighted by a ratio of 7:3), match the optimal adjustment threshold in the preset adjustment threshold table. The preset adjustment threshold table is generated through orthogonal experimental design and covers the thresholds corresponding to different combinations of the difference in fluctuation density, temporal distance, and peak difference. For example, when the linear combination value is 0.05, look up the table to obtain the optimal adjustment threshold of 46°C. This threshold is used to define whether the current temperature fluctuation needs adjustment and the adjustment amplitude.

[0072] Taking an industrial boiler as an example, its temperature regulation range is [40°C, 60°C], and the current fluctuation density difference is 0.22. The system retrieves the closest regulation mode from historical data with a difference of 0.23. The corresponding historical regulation threshold for this mode is 48°C, and the time series distribution shows a fluctuation peak every two hours. The current data shows a peak every hour. When calculating the Euclidean distance of the time series distribution, the 24-hour data is divided into 24 time points. The difference at each point is the difference in the fluctuation density between the current and the historical mode. After taking the square root of the sum of squares, the first threshold reference value is 0.08; the difference in the number of peaks is 12 (current) - 6 (historical) = 6, which is used as the second threshold reference value. The value after linear combination is 0.08×0.7 + 6×0.3 = 0.056 + 1.8 = 1.856. Looking up the table and matching, the optimal regulation threshold is 47°C.

[0073] When the regulation execution module analyzes the threshold deviation sequence, it first divides the positive fluctuation domain and the negative fluctuation domain according to the temperature fluctuation trend of each node. The positive fluctuation domain refers to the set of nodes where the temperature shows an upward trend. For example, for a node sequence [42°C, 44°C, 46°C], the difference between adjacent nodes is positive; the negative fluctuation domain refers to the set of nodes where the temperature shows a downward trend, such as [48°C, 45°C, 43°C], and the differences are all negative. In the positive fluctuation domain, calculate the convergence frequency of the threshold deviation, that is, the number of times the deviation value (current temperature - regulation threshold) shrinks per unit time; in the negative fluctuation domain, calculate the diffusion frequency of the threshold deviation, that is, the number of times the deviation value expands.

[0074] For example, in the positive fluctuation domain, the deviation of a node is +3°C (current 48°C, threshold 45°C), the deviation at the next moment is +2°C, and the next moment is +1°C, then the convergence frequency is 2 times per unit time; in the negative fluctuation domain, the deviation of a node is -2°C (current 43°C, threshold 45°C), the next moment is -3°C, and the next moment is -4°C, then the diffusion frequency is 2 times per unit time. Dynamically reconcile the convergence frequency and the diffusion frequency according to the regulation weight of the temperature node. The regulation weight is preset according to the thermal sensitivity of each component of the device. For example, the weight of the node near the heating element is 0.6, and the weight of the node in the heat dissipation area is 0.4.

[0075] Continuing with the boiler as an example, the threshold deviation sequence is as follows: Node 1 (46°C, deviation +1°C, positive), Node 2 (48°C, deviation +3°C, positive), Node 3 (47°C, deviation +2°C, positive), Node 4 (45°C, deviation 0°C, stable), Node 5 (43°C, deviation -2°C, negative), Node 6 (41°C, deviation -4°C, negative). The positive fluctuation domain is Nodes 1 - 3, and the convergence frequency is 1 time (the deviation from Node 2 to Node 3 shrinks from +3°C to +2°C); the negative fluctuation domain is Nodes 5 - 6, and the diffusion frequency is 1 time (the deviation from Node 5 to Node 6 expands from -2°C to -4°C). The adjustment weights for each node are: Node 1 (heating zone, 0.5), Node 2 (heating zone, 0.6), Node 3 (heating zone, 0.5), Node 5 (heat dissipation zone, 0.3), Node 6 (heat dissipation zone, 0.3).

[0076] During dynamic adjustment, first calculate the weighted convergence frequency of the positive domain: (1 time × 0.5 + 1 time × 0.6 + 0 time × 0.5) / (0.5 + 0.6 + 0.5) = 1.1 / 1.6 ≈ 0.6875 times per unit time; the weighted diffusion frequency of the negative domain: (0 time × 0.3 + 1 time × 0.3) / (0.3 + 0.3) = 0.3 / 0.6 = 0.5 times per unit time. The two are summed according to the preset weights (such as 0.7 for the positive proportion and 0.3 for the negative proportion) to obtain the comprehensive adjustment parameter: 0.6875 × 0.7 + 0.5 × 0.3 = 0.48125 + 0.15 = 0.63125. An execution plan is generated based on this parameter: Start the cooling fan in the positive domain and reduce the heating power by 10%; in the negative domain, increase the fuel supply, increase the heating power by 15%, and at the same time close some of the heat dissipation dampers.

[0077] In practical applications, the update frequency of historical data needs to match the equipment operation cycle. For example, automatically synchronize the latest 24 - hour data at midnight every day to ensure the timeliness of the adjustment mode. The maintenance of the preset adjustment threshold table needs to be combined with the equipment maintenance record. When key components of the equipment (such as heat exchangers) are replaced, orthogonal tests are re - conducted through manual intervention to update the parameters in the table. The setting of the adjustment weight needs to consider the heat conduction path of the equipment. For example, the weight of nodes far from the heat source is lower, and the weight of nodes close to the sensor is higher to give priority to ensuring the temperature accuracy of the measurement points.

[0078] The entire process uses the similarity matching of historical data and multi - dimensional parameter analysis to transform the abstract fluctuation characteristics into specific adjustment thresholds and execution parameters, achieving a closed - loop control from data to actions. The differential processing of the positive and negative fluctuation domains, combined with the dynamic adjustment of node weights, ensures the pertinence and effectiveness of the adjustment plan, enabling the system to quickly make accurate responses in complex temperature fluctuation scenarios and ultimately achieving the high - precision temperature control goal of the industrial thermostat.

[0079] It should be noted that in this text, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements not only includes those elements, but also includes other elements not expressly listed, or elements inherent to such process, method, article or device.

[0080] Although the embodiments of the present invention have been shown and described, it will be understood by those of ordinary skill in the art that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present invention, and the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An industrial thermostat temperature regulation system with high precision and fast response, characterized in that, Including: A temperature acquisition module, which is used to obtain the real-time temperature data of the target device and the environmental thermal inertia parameters, and set a temperature adjustment range that matches the device operation state. The temperature adjustment range is the temperature fluctuation data to be processed; A fluctuation analysis module, which is used to divide multiple temperature nodes within the temperature adjustment range, perform pattern recognition on the fluctuation data of each temperature node, and generate a fluctuation feature vector corresponding to the temperature node; A dynamic calibration module, which is used to extract temperature deviation indicators from the fluctuation feature vector, establish an adjustment compensation rule associated with the temperature node, and obtain the threshold control parameters corresponding to the adjustment compensation rule; A gradient optimization module, which is used to identify the temperature change level in the threshold control parameters, dynamically compensate the temperature deviation indicators according to the temperature change level, and calculate the fluctuation density difference of each temperature node under different adjustment strategies; A threshold management module, which is used to deduce the optimal adjustment threshold according to the fluctuation density difference, and generate a threshold deviation sequence by comparing the current temperature fluctuation density with the optimal adjustment threshold; An adjustment execution module, which is used to analyze the threshold deviation sequence, and integrate the threshold deviation sequence into a temperature adjustment execution plan based on the fluctuation density characteristics of the temperature node.

2. The high-precision and fast-response industrial thermostat temperature regulation system according to claim 1, characterized in that The implementation method of the fluctuation analysis module includes: constructing a device feature library corresponding to the temperature node, and the device feature library includes a temperature parameter vector mapped by real-time temperature data and environmental thermal inertia parameters; Performing similar fluctuation matching on the temperature parameter vector, and dividing the fluctuation clustering group of the temperature parameter vector according to the matching result; extracting the distribution balance point of the temperature data from the fluctuation clustering group, and setting the distribution balance point as the temperature node.

3. The high-precision and fast-response industrial thermostat temperature regulation system according to claim 2, characterized in that, Dividing the fluctuation clustering group of the temperature parameter vector further includes: Extracting the temperature change rate, adjustment interval and thermal inertia intensity parameters according to the temperature control frequency and thermal inertia fluctuation index in the temperature parameter vector, and generating a fluctuation feature label based on the temperature change rate, adjustment interval and thermal inertia intensity parameters; Associating the fluctuation feature label with the temperature parameter vector, and screening the parameter vectors with a fluctuation similarity higher than the preset fluctuation threshold to form a fluctuation clustering group by calculating the fluctuation similarity between the feature labels.

4. The high-precision and fast-response industrial thermostat temperature regulation system according to claim 1, characterized in that, The implementation method of generating a fluctuation feature vector corresponding to the temperature node includes: For each temperature node, according to the time sequence distribution of the temperature node in the temperature adjustment range, obtaining the fluctuation difference parameter of the temperature node within a preset period, and calculating the temperature difference coefficient of the node; When the temperature difference coefficient exceeds the first calibration threshold, marking the node as an unbalanced node, and extracting its temperature data to form a fluctuation feature vector; when the temperature difference coefficient is lower than the first calibration threshold, marking the node as a stable node, and performing gradient superposition on the temperature data of the adjacent nodes of the node, and reconstructing the superimposed data into a fluctuation feature vector.

5. The high-precision and fast-response industrial thermostat temperature regulation system according to claim 1, wherein, The implementation method of the dynamic calibration module includes: Separating the temperature data ratio, abnormal fluctuation ratio and thermal inertia fluctuation parameters from the fluctuation feature vector, and generating an adjustment compensation rule for the temperature node based on the above parameters; If the number of temperature nodes covered by the current adjustment compensation rule is less than the preset fluctuation threshold, traversing the fluctuation feature vectors of adjacent temperature nodes, and adding the temperature indicators not included in the adjustment compensation rules of the adjacent nodes to the current rule.

6. The high-precision and fast-response industrial thermostat temperature regulation system according to claim 1, characterized in that, The implementation methods of the gradient optimization module include: obtaining the timing parameters of the adjustment validity period and the fluctuation parameters of the gradient compensation intensity in the temperature change level; Constructing a compensation state network associated with the timing parameters and the fluctuation parameters, and determining the fluctuation density difference under different adjustment strategies according to the switching probabilities of various paths in the compensation state network.

7. The high-precision and fast-response industrial thermostat temperature regulation system according to claim 6, characterized in that, Constructing the compensation state network also includes: Identifying the periodic pattern of the timing parameters. If the current periodic pattern exactly matches the preset temperature control period, setting the timing parameters as the starting point of the compensation state network; Calculating the gradient correlation degree between the timing parameters and the fluctuation parameters, and generating the intermediate points and the end points of the compensation state network in descending order of the gradient correlation degree; Performing a state reverse verification on the end point. When the gradient correlation degree of the end point is lower than the preset gradient threshold, outputting it as the final link of the compensation state network.

8. The high-precision and fast-response industrial thermostat temperature regulation system according to claim 7, characterized in that, The implementation methods of calculating the fluctuation density difference include: Statistically analyzing the mean value of the timing parameters and the range of the fluctuation parameters of each end point in the compensation state network, and calculating the global covariance of all node parameters; Taking the difference between the mean value of the timing parameters of a single end point and the mean value of the timing parameters of the adjacent node, dividing the obtained difference by the global covariance to obtain the timing difference coefficient; at the same time, calculating the ratio of the range of the fluctuation parameters to the global covariance, and weighted summing the ratio and the timing difference coefficient as the fluctuation density difference of this node.

9. The high-precision and fast-response industrial thermostat temperature regulation system according to claim 1, characterized in that, The implementation methods of deriving the optimal adjustment threshold include: Extracting the adjustment mode in the historical data that is closest to the current fluctuation density difference, and calculating the Euclidean distance between the fluctuation density difference in the closest adjustment mode and the current fluctuation density difference in the timing distribution as the first threshold reference value; Statistically analyzing the difference in the number of peak points between the fluctuation density difference in the closest adjustment mode and the current fluctuation density difference, and taking the difference quantity as the second threshold reference value; Based on the linear combination of the first threshold reference value and the second threshold reference value, matching the optimal adjustment threshold in the preset adjustment threshold table.

10. The high-precision and fast-response industrial thermostat temperature regulation system according to claim 1, wherein, The implementation methods of the adjustment execution module include: dividing the positive fluctuation domain and the negative fluctuation domain according to the temperature fluctuation trend of each node in the threshold deviation sequence; Extracting the convergence frequency of the threshold deviation in the positive fluctuation domain and the diffusion frequency of the threshold deviation in the negative fluctuation domain, and dynamically adjusting the two according to the adjustment weight of the temperature node to generate the configuration parameters of the temperature adjustment execution plan.

Images