An industrial temperature intelligent monitoring system and method for the Internet of Things

Abstract

The present invention provides an industrial temperature intelligent monitoring system and method for the Internet of Things. The system includes a deviation weight allocation module, a deviation weighting module, a controllability evaluation module, a first initialization module, and a second initialization module. Among them, the deviation weight allocation module is used to obtain the weight time series value of the monitored temperature; the deviation weighting module is used to obtain the temperature deviation coefficient; the controllability evaluation module is used to obtain the temperature controllability coefficient; the first initialization module is used to initialize the temperature controller when the temperature controllability coefficient is greater than or equal to the temperature controllability coefficient threshold; the second initialization module is used to initialize the temperature controller when the temperature controllability coefficient is less than the temperature controllability coefficient threshold. Through deviation weight allocation and controllability evaluation, it is possible to intelligently judge the influence degree of temperature fluctuations and selectively optimize control parameters, improving the refinement degree of industrial temperature monitoring.

Description

Technical Field

[0001] The present invention relates to the field of temperature control, and particularly to an industrial temperature intelligent monitoring system and method for the Internet of Things. Background Art

[0002] As one of the key parameters affecting the quality of industrial products, accurate control of temperature is of great significance for ensuring product quality and improving production efficiency. At present, most industrial temperature monitoring adopts predictive control methods, that is, by predicting the temperature change trend of products, when the prediction result does not meet the production requirements, the adjustment process of temperature control parameters is immediately started. However, this temperature monitoring method often responds to any fluctuations in temperature influencing factors. Regardless of the magnitude of the impact of the fluctuations, it will trigger the adjustment process of temperature control parameters. For example, for temperature fluctuations with a small impact degree, a complete parameter adjustment process will also be executed, resulting in waste of resources and unnecessary intervention, and insufficient refinement of temperature monitoring. Summary of the Invention

[0003] Aiming at the technical problem in the prior art that industrial temperature monitoring immediately adjusts temperature parameters only based on the fluctuations of influencing factors without considering the impact degree of the fluctuations, resulting in insufficient refinement of temperature monitoring, the present invention provides an industrial temperature intelligent monitoring system and method for the Internet of Things to solve this problem.

[0004] The technical solution of the present invention to solve the above technical problems is as follows:

[0005] In a first aspect, the present invention provides an industrial temperature intelligent monitoring system for the Internet of Things, including: a deviation weight distribution module, configured to receive the time series value of the monitored environmental temperature and the pre-stored environmental temperature value for deviation weight distribution to obtain the time series value of the monitored temperature weight; a deviation weighting module, configured to weight the deviation between the time series value of the monitored environmental temperature and the pre-stored environmental temperature value according to the time series value of the monitored temperature weight to obtain a temperature deviation coefficient; a controllability evaluation module, configured to retrieve, with the target product model and production control parameters as constraints, the proportion of samples whose temperature fluctuations are less than the temperature fluctuation threshold among the samples with the same constraints that meet the temperature deviation coefficient, which is set as the temperature controllability coefficient; a first initialization module, configured to initialize the temperature controller according to the preset temperature control parameters when the temperature controllability coefficient is greater than or equal to the temperature controllability coefficient threshold; a second initialization module, configured to optimize the preset temperature control parameters based on the time series value of the monitored environmental temperature to obtain the target temperature control parameters for initializing the temperature controller when the temperature controllability coefficient is less than the temperature controllability coefficient threshold, and update the pre-stored environmental temperature value with the centralized temperature value of the time series value of the monitored environmental temperature at the same time.

[0006] Optionally, the deviation weight allocation module includes: a temperature clustering unit for performing neighborhood hierarchical clustering on the environmental monitoring temperature time series values to obtain an environmental temperature clustering value sequence; a deviation calculation unit for traversing the environmental temperature clustering value sequence and calculating a temperature deviation value sequence from the pre-stored environmental temperature values; and a weight allocation unit for traversing the temperature deviation value sequence and obtaining the monitoring temperature weight time series values by taking the ratio to the sum of the temperature deviations of the temperature deviation value sequence.

[0007] Optionally, the controllability evaluation module includes: a constraint condition construction unit for constructing query constraint conditions based on the target product model and the production control parameters; a sample retrieval unit for retrieving the pre-stored environmental temperature record values, environmental temperature monitoring record time series values, product temperature pre-stored record time series values, and product temperature monitoring record time series values of the first same-constraint samples that meet the query constraint conditions; a sample deviation analysis unit for calculating the sample temperature deviation coefficient between the pre-stored environmental temperature record values and the environmental temperature monitoring record time series values; a temperature fluctuation analysis unit for calculating the proportion of time series with a temperature deviation greater than or equal to the temperature deviation threshold between the pre-stored product temperature record time series values and the product temperature monitoring record time series values, set as the first temperature fluctuation value, when the deviation between the sample temperature deviation coefficient and the temperature deviation coefficient is less than or equal to the temperature deviation coefficient tolerance threshold; until the Qth temperature fluctuation value is obtained, where Q is an integer and Q≥500; and a sample statistics unit for statistically calculating the proportion of samples with a temperature fluctuation value less than the temperature fluctuation threshold from the first temperature fluctuation value to the Qth temperature fluctuation value, set as the temperature controllability coefficient.

[0008] Optionally, the controllability evaluation module further includes: a sample supplement trigger unit for obtaining the sample production control parameters of the first same-constraint samples when the number of the first same-constraint samples that meet the query constraint conditions is less than 2Q; a constraint update unit for constructing query constraint update conditions based on the target product model and the sample production control parameters; a supplementary sample retrieval unit for retrieving the second same-constraint samples that meet the query constraint update conditions; and a comprehensive calculation unit for calculating the temperature controllability coefficient by combining the first same-constraint samples and the second same-constraint samples.

[0009] Optionally, the second initialization module includes: a predictor acquisition unit configured to obtain a product temperature predictor bound to the target product model and the temperature controller model; a centralized temperature statistics unit configured to statistically calculate the centralized temperature value of the ambient monitoring temperature time series values; a parameter update unit configured to update the preset temperature control parameters to obtain temperature update control parameters; a temperature prediction unit configured to input the centralized temperature value and the temperature update control parameters into the product temperature predictor and output product temperature prediction time series information; and a parameter confirmation unit configured to set the temperature update control parameters as the temperature control parameters when the product temperature prediction time series information is consistent with the product reference temperature time series information.

[0010] Optionally, the predictor acquisition unit includes: a data acquisition sub-unit configured to acquire multiple sets of data with the target product model and the temperature controller model as static constraints, with the ambient temperature and the temperature control parameters as independent variables, and with the product temperature as the dependent variable; a first function construction unit configured to construct a first loss function, wherein the first loss function is used to statistically calculate the proportion of moments when the temperature deviation between the product temperature supervision time series information and the product temperature prediction time series information is greater than or equal to a temperature deviation threshold, and the first loss function has a first weight; a second function construction unit configured to construct a second loss function, wherein the second loss function is used to statistically calculate the average temperature deviation between the product temperature supervision time series information and the product temperature prediction time series information, and the second loss function has a second weight; and a predictor training sub-unit configured to add the first loss function and the second loss function according to the first weight and the second weight to obtain a loss function, retrieve the multiple sets of data, and train the product temperature predictor.

[0011] Optionally, the second initialization module further includes: a parameter iteration unit, configured to execute a loop for adjusting the temperature update control parameter when the product temperature prediction time series information is inconsistent with the product reference temperature time series information; a parameter set statistics unit, configured to obtain an analyzed temperature control parameter set until the number of loops meets a preset number of loops; a dynamic distance evaluation unit, configured to traverse the product temperature prediction time series information set of the analyzed temperature control parameter set, and perform a dynamic time warping distance evaluation with the product reference temperature time series information to obtain a dynamic time warping distance set; a parameter sorting unit, configured to sort the analyzed temperature control parameter set in ascending order according to the dynamic time warping distance set to obtain an analyzed temperature control parameter set sequence; a preferred parameter extraction unit, configured to extract the top k analyzed temperature control parameters in the sorted analyzed temperature control parameter set sequence, perform a mean calculation of the same attribute parameters, and obtain an adjusted end temperature control parameter; a parameter expansion unit, configured to adjust the 20% of the analyzed temperature control parameters with lower sorting positions in the analyzed temperature control parameter set sequence under the guidance of the adjusted end temperature control parameter to obtain an expanded temperature control parameter set and execute a loop.

[0012] In a second aspect, the present invention provides an industrial temperature intelligent monitoring method for the Internet of Things, including: receiving the environmental monitoring temperature time series value and the environmental temperature pre-stored value to perform deviation weight allocation to obtain a monitored temperature weight time series value; according to the monitored temperature weight time series value, weighting the deviation between the environmental monitoring temperature time series value and the environmental temperature pre-stored value to obtain a temperature deviation coefficient; taking the target product model and production control parameters as constraints, retrieving the proportion of samples with a temperature fluctuation less than a temperature fluctuation threshold among the same-constraint samples that satisfy the temperature deviation coefficient, and setting it as a temperature controllable coefficient; when the temperature controllable coefficient is greater than or equal to a temperature controllable coefficient threshold, initializing a temperature controller according to a preset temperature control parameter; when the temperature controllable coefficient is less than the temperature controllable coefficient threshold, optimizing the preset temperature control parameter based on the environmental monitoring temperature time series value to obtain a target temperature control parameter for initializing the temperature controller, and simultaneously updating the environmental temperature pre-stored value with the centralized temperature value of the environmental monitoring temperature time series value.

[0013] The beneficial effects of the present invention are:

[0014] Through the deviation weight allocation module, receive the environmental monitoring temperature time series value and the pre-stored environmental temperature value, conduct deviation weight allocation, and obtain the monitoring temperature weight time series value, which can weight the temperature deviation at different time points and highlight the temperature change characteristics of important periods. Through the deviation weighting module, weight the deviation between the environmental monitoring temperature time series value and the pre-stored environmental temperature value according to the monitoring temperature weight time series value to obtain the temperature deviation coefficient, providing a quantitative basis for subsequent decision-making. Through the controllability evaluation module, with the target product model and production control parameters as constraints, retrieve the proportion of samples with temperature fluctuations less than the temperature fluctuation threshold among the same-constraint samples that meet the temperature deviation coefficient, set as the temperature controllability coefficient, evaluate the controllability under the current temperature deviation, and provide a decision-making basis for how to adjust the temperature control parameters.

[0015] When the temperature controllability coefficient is greater than or equal to the temperature controllability coefficient threshold, initialize the temperature controller according to the preset temperature control parameters through the first initialization module, indicating that the current temperature fluctuation is within the controllable range and there is no need to specially adjust the control parameters. When the temperature controllability coefficient is less than the temperature controllability coefficient threshold, optimize the preset temperature control parameters based on the environmental monitoring temperature time series value through the second initialization module to obtain the target temperature control parameters to initialize the temperature controller, and at the same time update the pre-stored environmental temperature value with the centralized temperature value of the environmental monitoring temperature time series value, so that when the temperature fluctuation exceeds the controllable range, by optimizing and adjusting the control parameters and updating the temperature reference value, the adaptive adjustment of the temperature is realized.

[0016] Through the above technical solutions, this application can intelligently judge the influence degree of temperature fluctuations, perform parameter optimization only when necessary, avoid unnecessary adjustments, and improve the refinement degree and efficiency of industrial temperature monitoring. Description of the Drawings

[0017] Figure 1 It is a schematic structural diagram of an industrial temperature intelligent monitoring system of the Internet of Things provided by the present invention;

[0018] Figure 2 It is a schematic flow diagram of an industrial temperature intelligent monitoring method of the Internet of Things provided by the present invention.

[0019] In the drawings, the components represented by each reference numeral are as follows:

[0020] Deviation weight allocation module 11, deviation weighting module 12, controllability evaluation module 13, first initialization unit 14, second initialization unit 15. Detailed Embodiment

[0021] The following will clearly and completely describe the technical solutions in the embodiments of the present invention 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. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative efforts belong to the scope of protection of the present invention.

[0022] In the description of the present invention, the terms "first" and "second" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include one or more of the described features. In the description of the present invention, "a plurality" means two or more, unless otherwise specifically defined.

[0023] In the description of the present invention, the term "for example" is used to mean "serving as an example, illustration, or explanation". Any embodiment described as "for example" in the present invention is not necessarily construed as being more preferred or advantageous than other embodiments. The following description is provided to enable any person skilled in the art to implement and use the present invention. Details are set forth for the purpose of explanation in the following description. It should be understood that those of ordinary skill in the art can recognize that the present invention can be implemented without the use of these specific details. In other instances, well-known structures and processes are not described in detail to avoid obscuring the description of the present invention with unnecessary details. Therefore, the present invention is not intended to be limited to the embodiments shown, but is to be accorded the widest scope consistent with the principles and features disclosed herein.

[0024] Embodiment 1, as Figure 1 shown, the embodiment of the present invention provides an industrial temperature intelligent monitoring system for the Internet of Things, which includes a deviation weight allocation module 11, a deviation weighting module 12, a controllability evaluation module 13, a first initialization module 14, and a second initialization module 15.

[0025] The deviation weight allocation module 11 is used to receive the environmental monitoring temperature time series value and the pre-stored environmental temperature value for deviation weight allocation to obtain the monitoring temperature weight time series value.

[0026] Specifically, the function of the deviation weight allocation module 11 is to receive the environmental monitoring temperature time series value and compare it with the pre-stored environmental temperature value, so as to allocate the corresponding deviation weight.

[0027] First, the deviation weight allocation module 11 obtains the real-time collected environmental monitoring temperature time series value, which characterizes the change of temperature over time in the industrial production environment. Subsequently, the deviation weight allocation module 11 compares and analyzes the environmental monitoring temperature time series value with the pre-stored environmental temperature pre-stored value (i.e., the preset standard environmental temperature value considered to be the most suitable), and assigns corresponding weight coefficients to the environmental monitoring temperature at each time point according to the deviation magnitude and deviation distribution characteristics between the two, obtaining the monitored temperature weight time series value. This monitored temperature weight time series value reflects the importance of the environmental temperature deviation at each time point and provides a refined basis for subsequent temperature control decisions.

[0028] The deviation weighting module 12 is used to weight the deviation between the environmental monitoring temperature time series value and the environmental temperature pre-stored value according to the monitored temperature weight time series value to obtain the temperature deviation coefficient.

[0029] Specifically, first, the deviation weighting module 12 receives the monitored temperature weight time series value output from the deviation weight allocation module 11, which represents the importance of temperature deviation at different time points. Subsequently, the deviation weighting module 12 calculates the specific deviation value between the environmental monitoring temperature time series value and the environmental temperature pre-stored value at each time point. Then, the deviation weighting module 12 multiplies the above deviation value by the monitored temperature weight at the corresponding time point to achieve the weighting process of the deviation, and calculates the mean value of all weighted deviation values to output the temperature deviation coefficient. This temperature deviation coefficient is a comprehensive quantitative index of the impact degree of environmental temperature fluctuation on the production process and provides parameters for subsequent evaluation of temperature controllability.

[0030] Through the weighting process, it can more accurately reflect the actual impact of environmental temperature fluctuation on the production process, thus avoiding the problem of insufficient evaluation of the impact degree of temperature fluctuation in the traditional monitoring system.

[0031] The controllability evaluation module 13 is used to retrieve the proportion of samples with temperature fluctuation less than the temperature fluctuation threshold among the same-constraint samples that meet the temperature deviation coefficient, with the target product model and production control parameters as constraints, and set it as the temperature controllability coefficient.

[0032] Specifically, the function of the controllability evaluation module 13 is to evaluate the possibility and effectiveness of controlling the temperature under the current temperature deviation situation, so as to form a quantitative temperature controllability coefficient.

[0033] First, the controllability evaluation module 13 uses the target product model and production control parameters as constraints to determine the specific conditions and control requirements of the current production environment. Subsequently, based on these constraints, the controllability evaluation module 13 retrieves from the historical database the same-constraint samples that match the current situation. Among these samples, the controllability evaluation module 13 filters out those samples whose temperature deviation coefficients are close to or the same as the temperature deviation coefficient currently calculated by the deviation weighting module 12. After that, the controllability evaluation module 13 analyzes these filtered samples, calculates the proportion of the number of samples with temperature fluctuations less than the predetermined temperature fluctuation threshold in the total number of samples, and sets this proportion as the temperature controllability coefficient. This temperature controllability coefficient is a value between 0 and 1, reflecting the likelihood of effectively controlling the product temperature under the current temperature deviation conditions. The higher the temperature controllability coefficient, the more likely it is to achieve effective temperature control under the current conditions; conversely, it indicates greater difficulty in temperature control.

[0034] The first initialization module 14 is used to initialize the temperature controller according to the preset temperature control parameters when the temperature controllability coefficient is greater than or equal to the temperature controllability coefficient threshold.

[0035] Specifically, the function of the first initialization module 14 is to initialize the temperature controller with the preset temperature control parameters based on the temperature controllability evaluation result (i.e., the temperature controllability coefficient) when it is determined that there is sufficient temperature control ability.

[0036] First, the first initialization module 14 receives the temperature controllability coefficient calculated by the controllability evaluation module 13 and compares this temperature controllability coefficient with the preset temperature controllability coefficient threshold. When the temperature controllability coefficient is greater than or equal to the temperature controllability coefficient threshold, it indicates that under the current ambient temperature deviation conditions, the product temperature fluctuation can be effectively controlled and the temperature stability of the production process can be maintained. In this case, the first initialization module 14 determines to adopt the conventional control method without special adjustment of the control parameters. Subsequently, the first initialization module 14 directly calls the preset temperature control parameters. The preset temperature control parameters are standard parameter values optimized based on a large amount of historical production data and are applicable to most normal production situations. After that, the first initialization module 14 initializes the temperature controller with the preset temperature control parameters, enabling the temperature controller to operate according to the preset standard control strategy, thereby ensuring stable and reliable temperature control during the product production process.

[0037] The second initialization module 15 is used to optimize the preset temperature control parameters based on the environmental monitoring temperature time series value to obtain the target temperature control parameters for initializing the temperature controller when the temperature controllability coefficient is less than the temperature controllability coefficient threshold, and at the same time update the pre-stored environmental temperature value with the central temperature value of the environmental monitoring temperature time series value.

[0038] Specifically, the function of the second initialization module 15 is to optimize and adjust the temperature control parameters and update the ambient temperature reference value when the temperature control ability is insufficient, so as to improve the adaptability to abnormal temperature fluctuations.

[0039] First, the second initialization module 15 receives the temperature controllability coefficient calculated by the controllability evaluation module 13 and compares the temperature controllability coefficient with a preset temperature controllability coefficient threshold. When the temperature controllability coefficient is less than the temperature controllability coefficient threshold, it indicates that it is difficult to effectively control the ambient temperature using conventional preset temperature control parameters under the current ambient temperature deviation conditions, and there is a high risk of temperature runaway. In this case, the second initialization module 15 starts the parameter optimization process, and based on the currently collected ambient monitoring temperature time series values, pertinently optimizes and adjusts the preset temperature control parameters. Through optimization, the second initialization module 15 obtains the target temperature control parameters adapted to the current ambient temperature fluctuation characteristics and initializes the temperature controller using the optimized target temperature control parameters. At the same time, the second initialization module 15 analyzes the distribution characteristics of the ambient monitoring temperature time series values, extracts its centralized temperature value (such as the median or weighted average), and updates the ambient temperature pre-stored value with the centralized temperature value, so that the reference temperature value can dynamically adapt to the change of the ambient temperature and provide a more accurate reference benchmark for subsequent temperature monitoring and control.

[0040] Through the collaborative work of the deviation weight distribution module 11, the deviation weighting module 12, the controllability evaluation module 13, the first initialization module 14, and the second initialization module 15, the refined monitoring and intelligent control of the industrial production environment temperature change are realized. In the embodiment of the present application, first, the ambient temperature deviation is weighted and weighted to obtain a temperature deviation coefficient; then, the temperature controllability is evaluated based on historical samples to form a temperature controllability coefficient; after that, according to the temperature controllability coefficient, corresponding controller initialization strategies are adopted. Compared with traditional industrial temperature monitoring, the embodiment of the present application not only considers the existence of temperature fluctuations, but also pays more attention to the degree of influence of the fluctuations. By controllability evaluation, unnecessary parameter adjustments are avoided, and targeted optimization is carried out when necessary, improving the accuracy, adaptability, and efficiency of temperature control, reducing the product quality fluctuations caused by temperature fluctuations during the production process, and providing a more stable and reliable temperature environment guarantee for industrial production.

[0041] Further, the deviation weight distribution module 11 includes:

[0042] A temperature clustering unit, configured to perform neighborhood hierarchical clustering on the ambient monitoring temperature time series values to obtain an ambient temperature clustering value sequence;

[0043] A deviation calculation unit, configured to traverse the ambient temperature clustering value sequence and calculate a temperature deviation value sequence from the ambient temperature pre-stored value;

[0044] A weight allocation unit, configured to traverse the temperature deviation value sequence and calculate the ratio with the sum of temperature deviations of the temperature deviation value sequence, so as to obtain the monitored temperature weight time sequence value.

[0045] In an optional implementation manner, the deviation weight allocation module 11 includes a temperature clustering unit, a deviation calculation unit, and a weight allocation unit. These units work together to achieve refined processing and weight allocation of ambient temperature time series data.

[0046] First, the temperature clustering unit receives the ambient temperature monitoring time sequence value and performs neighborhood hierarchical clustering processing on it. Neighborhood hierarchical clustering is a bottom-up clustering method that can group similar temperature values into one category according to the distance relationship between temperature data points. Through neighborhood hierarchical clustering, the temperature clustering unit can effectively identify the main temperature characteristics and data patterns in the ambient temperature monitoring time sequence value, filter out the influence of random noise, and output the ambient temperature clustering value sequence. This ambient temperature clustering value sequence represents the main distribution characteristics of the ambient temperature and provides a more stable and reliable data basis for subsequent analysis.

[0047] Subsequently, the deviation calculation unit traverses each clustering value in the above ambient temperature clustering value sequence, compares it with the pre-stored ambient temperature value, calculates the difference between the two, and thus obtains the temperature deviation value sequence. This temperature deviation value sequence reflects the deviation degree of the ambient temperature relative to the pre-stored temperature, where each deviation value corresponds to the difference degree between a clustering value in the ambient temperature clustering value sequence and the pre-stored ambient temperature value.

[0048] After that, the weight allocation unit traverses each deviation value in the temperature deviation value sequence, divides the absolute value of this deviation value by the sum of deviations of the temperature deviation value sequence (i.e., the sum of the absolute values of all deviation values), and obtains the proportion of each deviation value. These proportion values constitute the monitored temperature weight time sequence value. Through this normalization process, the temperature deviations at different time points are converted into corresponding weight coefficients, realizing the quantitative evaluation and weight allocation of the ambient temperature deviation, and providing a basis for subsequent weighted processing.

[0049] Furthermore, the controllability evaluation module 13 includes:

[0050] A constraint condition construction unit, configured to construct query constraint conditions based on the target product model and the production control parameters;

[0051] A sample retrieval unit, configured to retrieve the pre-stored ambient temperature record value, the ambient temperature monitoring record time sequence value, the pre-stored product temperature record time sequence value, and the product temperature monitoring record time sequence value of the first same-constraint sample that satisfies the query constraint conditions;

[0052] A sample deviation analysis unit for calculating the sample temperature deviation coefficient of the pre - stored record value of the environmental temperature and the time - series value of the monitored environmental temperature record;

[0053] A temperature fluctuation analysis unit for calculating the proportion of time series in which the temperature deviation between the pre - stored record time - series value of the product temperature and the monitored record time - series value of the product temperature is greater than or equal to the temperature deviation threshold, set as the first temperature fluctuation value, when the deviation between the sample temperature deviation coefficient and the temperature deviation coefficient is less than or equal to the temperature deviation coefficient tolerance threshold; until the Q - th temperature fluctuation value is obtained, where Q is an integer and Q≥500;

[0054] A sample statistics unit for statistically calculating the proportion of samples less than the temperature fluctuation threshold among the first temperature fluctuation value to the Q - th temperature fluctuation value, set as the temperature controllability coefficient.

[0055] In a preferred embodiment, the controllability evaluation module 13 includes a constraint condition construction unit, a sample retrieval unit, a sample deviation analysis unit, a temperature fluctuation analysis unit, and a sample statistics unit to achieve an accurate evaluation of temperature controllability based on historical data.

[0056] First, the constraint condition construction unit constructs query constraint conditions based on the target product model and production control parameters in the current production environment. The query constraint conditions determine the historical sample screening criteria that match the current production situation, ensuring that the samples retrieved subsequently have a high similarity to the current production conditions. Subsequently, the sample retrieval unit retrieves samples that meet the query constraint conditions from the historical database according to the constructed query constraint conditions above, which are the first same - constraint samples. For each sample in the first same - constraint samples, the sample retrieval unit extracts its pre - stored record value of the environmental temperature, the time - series value of the monitored environmental temperature record, the pre - stored record time - series value of the product temperature, and the time - series value of the monitored product temperature record. These record values contain the complete change information of the environmental temperature and the product temperature in the historical production process, providing a data basis for subsequent analysis.

[0057] Next, the sample deviation analysis unit processes each sample in the first same-constraint sample, and calculates the sample temperature deviation coefficient between its pre-stored environmental temperature record value and the time series value of the monitored environmental temperature record. This sample temperature deviation coefficient characterizes the fluctuation of the current time series value of the monitored environmental temperature relative to the pre-stored environmental temperature record value in the historical samples, and is used to compare with the current temperature deviation coefficient. Then, the temperature fluctuation analysis unit filters out the samples whose difference between the sample temperature deviation coefficient and the current temperature deviation coefficient (calculated by the deviation weighting module 12) is less than or equal to the temperature deviation coefficient tolerance threshold. For these samples, the temperature fluctuation analysis unit calculates the proportion of the time points when the temperature deviation between the pre-stored time series value of the product temperature and the monitored time series value of the product temperature is greater than or equal to the temperature deviation threshold, and sets this proportion as the first temperature fluctuation value. The temperature fluctuation analysis unit repeats the above process until at least Q temperature fluctuation values are obtained, where Q is an integer not less than 500, ensuring the statistical reliability of the evaluation result. Specifically, first, the temperature fluctuation analysis unit obtains the pre-stored time series value of the product temperature (i.e., the target product temperature sequence in historical production) and the monitored time series value of the product temperature (i.e., the actually measured product temperature sequence in historical production); then, the temperature fluctuation analysis unit compares these two sets of data one by one, and calculates the deviation between the actual product temperature and the target product temperature at each time point; next, the temperature fluctuation analysis unit identifies all the time points where the deviation value is greater than or equal to the temperature deviation threshold, denoted as abnormal time points, and these time points represent that the product temperature has fluctuated beyond the allowable range; after that, the temperature fluctuation analysis unit calculates the proportion of these abnormal time points in the entire time series, that is, the number of abnormal time points divided by the total number of time points, and sets this proportion value as the temperature fluctuation value of the sample. This temperature fluctuation value intuitively reflects the instability of the product temperature in the historical samples under specific environmental temperature deviation conditions, and provides an important reference for evaluating the feasibility of temperature control under the current production conditions.

[0058] Subsequently, the sample statistics unit performs statistical analysis on the first temperature fluctuation value to the Qth temperature fluctuation value obtained, calculates the proportion of the number of samples less than the temperature fluctuation threshold in the total number of samples, and sets this proportion as the temperature controllability coefficient. This temperature controllability coefficient intuitively reflects the ability to control the product temperature within the expected range under the current environmental temperature deviation conditions, and provides a basis for the selection of subsequent control strategies.

[0059] Furthermore, the controllability evaluation module 13 further includes:

[0060] A sample supplement trigger unit, configured to obtain the sample production control parameters of the first same-constraint sample when the number of the first same-constraint samples that meet the query constraint conditions is less than 2Q;

[0061] A constraint update unit for constructing a query constraint update condition based on the target product model and the sample production control parameters;

[0062] A supplementary sample retrieval unit for retrieving second co-constraint samples that meet the query constraint update condition;

[0063] An integrated calculation unit for calculating the temperature controllability coefficient by combining the first co-constraint sample and the second co-constraint sample.

[0064] In an alternative embodiment, the controllability evaluation module 13 further includes a sample supplement trigger unit, a constraint update unit, a supplementary sample retrieval unit, and an integrated calculation unit to achieve intelligent sample expansion and comprehensive evaluation in the case of insufficient sample quantity.

[0065] First, the sample supplement trigger unit monitors the quantity of the first co-constraint samples that meet the query constraint conditions. When it detects that the quantity of the first co-constraint samples is less than 2Q (where Q is an integer not less than 500), the sample supplement trigger unit determines that the existing sample quantity is insufficient to ensure the reliability of statistical analysis, and then triggers the sample supplement process. In this case, the sample supplement trigger unit extracts the sample production control parameters from the obtained first co-constraint samples, and these parameters contain the production control information in the historical samples. Subsequently, the constraint update unit uses the target product model and the above-extracted sample production control parameters to construct a query constraint update condition. Compared with the original query constraint condition, the query constraint update condition retains the core constraint of the product model, but adopts the parameter characteristics from the first co-constraint samples in terms of production control parameters, thereby expanding the sample screening range while maintaining the sample relevance.

[0066] Next, the supplementary sample retrieval unit uses the above query constraint update condition to retrieve second co-constraint samples that meet the query constraint update condition from the historical database. Although these second co-constraint samples have some differences in production control parameters from the current production situation, they are consistent in terms of product model and key production control parameters, so they can be used as effective supplementary samples to expand the scale of the original samples. Then, the integrated calculation unit integrates the original first co-constraint samples and the newly retrieved second co-constraint samples, and calculates the temperature controllability coefficient based on this expanded sample.

[0067] Through the sample supplement and comprehensive evaluation mechanism, it is possible to still obtain a statistically significant temperature controllability evaluation result in the case of insufficient original samples, effectively improving the applicability and evaluation accuracy under various production conditions.

[0068] Furthermore, the second initialization module 15 includes:

[0069] A predictor acquisition unit for obtaining a product temperature predictor bound to a target product model and a temperature controller model;

[0070] A centralized temperature statistics unit for statistically analyzing the centralized temperature values of the environmental monitoring temperature time series values;

[0071] A parameter update unit for updating the preset temperature control parameters to obtain temperature update control parameters;

[0072] A temperature prediction unit for inputting the centralized temperature value and the temperature update control parameters into the product temperature predictor and outputting product temperature prediction time series information;

[0073] A parameter confirmation unit for setting the temperature update control parameters as the temperature control parameters when the product temperature prediction time series information is consistent with the product reference temperature time series information.

[0074] In an alternative embodiment, the second initialization module 15 includes a predictor acquisition unit, a centralized temperature statistics unit, a parameter update unit, a temperature prediction unit, and a parameter confirmation unit, realizing intelligent parameter optimization and controller initialization when the temperature controllability is insufficient.

[0075] First, the predictor acquisition unit obtains a product temperature predictor that matches the target product model and the temperature controller model in the current production environment. This product temperature predictor is a pre-trained model that can predict the change of product temperature according to the environmental temperature and control parameters, providing a simulation evaluation ability for subsequent parameter optimization. Subsequently, the centralized temperature statistics unit statistically analyzes the currently collected environmental monitoring temperature time series values and calculates its centralized temperature value. This centralized temperature value reflects the main distribution characteristics of the current environmental temperature and can be obtained by statistical methods such as median, mode, or weighted average, and is used to characterize the main level of the current environmental temperature. Then, the parameter update unit updates the preset temperature control parameters to obtain temperature update control parameters. This update process involves fine-tuning or recalculating the parameters, aiming to make the control parameters better adapt to the current environmental temperature conditions.

[0076] Then, the temperature prediction unit takes the above-obtained centralized temperature value and temperature update control parameter as inputs and sends them to the product temperature predictor. Through predictive calculation, product temperature prediction timing information is obtained, which characterizes the expected change of the product temperature over time under the current ambient temperature and update control parameter. After that, the parameter confirmation unit compares the above product temperature prediction timing information with the product reference temperature timing information (i.e., the ideal temperature change curve of the product). When the two are consistent or the difference is within the allowable range, it indicates that the updated control parameter can effectively control the product temperature. At this time, the parameter confirmation unit sets the temperature update control parameter as the temperature control parameter for initializing the temperature controller.

[0077] Through the parameter optimization and verification mechanism based on the prediction model, it is possible to find effective control parameters even when the ambient temperature fluctuates greatly and is difficult to control, ensuring the stability of the product temperature.

[0078] Furthermore, the predictor acquisition unit includes:

[0079] The data acquisition sub-unit is used to collect multiple groups of data with the target product model and the temperature controller model as static constraints, the ambient temperature and the temperature control parameter as independent variables, and the product temperature as the dependent variable.

[0080] The first function construction unit is used to construct a first loss function, where the first loss function is used to count the proportion of moments when the temperature deviation between the product temperature supervision timing information and the product temperature prediction timing information is greater than or equal to the temperature deviation threshold, and the first loss function has a first weight.

[0081] The second function construction unit is used to construct a second loss function, where the second loss function is used to count the average value of the temperature deviation between the product temperature supervision timing information and the product temperature prediction timing information, and the second loss function has a second weight.

[0082] The predictor training sub-unit is used to add the first loss function and the second loss function according to the first weight and the second weight to obtain a loss function, retrieve the multiple groups of data, and train the product temperature predictor.

[0083] Specifically, the predictor acquisition unit includes a data acquisition sub-unit, a first function construction unit, a second function construction unit, and a predictor training sub-unit, realizing the construction and training of the temperature predictor for a specific product and controller.

[0084] First, based on determining the target product model and temperature controller model as static constraint conditions, the data acquisition sub-unit collects multiple groups of data from the historical production process, using the ambient temperature and temperature control parameters as independent variables and the product temperature as the dependent variable. These data contain the actual changes in product temperature under different combinations of ambient temperature and control parameters, providing a complete sample set for the training of the product temperature predictor. The collected data has clear physical meanings and corresponding relationships, that is, how the ambient temperature and control parameters affect the final product temperature.

[0085] Subsequently, the first function construction unit constructs a first loss function, which focuses on the overall deviation distribution between the predicted temperature and the actual temperature. Specifically, the first loss function calculates the proportion of the number of moments when the temperature deviation between the product temperature supervised time series information (i.e., the actually measured product temperature sequence) and the product temperature predicted time series information (i.e., the predicted product temperature sequence) is greater than or equal to the temperature deviation threshold to the total number of moments. The first loss function focuses on the proportion of time points outside the allowable error range in the prediction results, ensuring that the model can reduce the occurrence frequency of prediction deviations. Corresponding to the first loss function, a first weight is assigned to reflect its importance in model training. At the same time, the second function construction unit constructs a second loss function, which focuses on the average deviation level between the predicted temperature and the actual temperature. Specifically, the second loss function calculates the mean value of the temperature deviations between the product temperature supervised time series information and the product temperature predicted time series information, reflecting the overall average error level of the model prediction. Corresponding to the second loss function, a second weight is assigned to reflect its importance in model training.

[0086] After that, the predictor training sub-unit weights and sums the first loss function and the second loss function according to the first weight and the second weight to form a comprehensive loss function. This loss function not only considers the control of temperature abnormal deviations but also takes into account the overall prediction accuracy, providing a comprehensive optimization objective for the training of the product temperature predictor. The predictor training sub-unit retrieves the multiple groups of data collected by the data acquisition sub-unit and trains the product temperature predictor based on the loss function, enabling it to accurately predict the product temperature changes under specific ambient temperatures and control parameters, providing a reliable simulation evaluation tool for subsequent parameter optimization.

[0087] Furthermore, the second initialization module 15 further includes:

[0088] A parameter iteration unit, configured to execute a loop to adjust the temperature update control parameter when the product temperature predicted time series information is inconsistent with the product reference temperature time series information;

[0089] A parameter set statistics unit, configured to obtain the analyzed temperature control parameter set until the number of loops meets the preset number of loops;

[0090] A dynamic distance evaluation unit is configured to traverse the product temperature prediction time series information set of the analyzed temperature control parameter set, perform a dynamic time warping distance evaluation with the product reference temperature time series information, and obtain a dynamic time warping distance set;

[0091] A parameter sorting unit is configured to sort the analyzed temperature control parameter set in ascending order according to the dynamic time warping distance set, and obtain an analyzed temperature control parameter set sequence;

[0092] A preferred parameter extraction unit is configured to extract the top k analyzed temperature control parameters in the sorted analyzed temperature control parameter set sequence, perform a mean calculation of the same-attribute parameters, and obtain an adjusted end-point temperature control parameter;

[0093] A parameter expansion unit is configured to take the adjusted end-point temperature control parameter as a guide, adjust the 20% of the analyzed temperature control parameters at the end of the sorted analyzed temperature control parameter set sequence, and obtain an expanded temperature control parameter set to execute a loop.

[0094] In a preferred embodiment, the second initialization module 15 further includes a parameter iteration unit, a parameter set statistics unit, a dynamic distance evaluation unit, a parameter sorting unit, a preferred parameter extraction unit, and a parameter expansion unit to implement an adaptive parameter optimization process when the product temperature prediction time series information is inconsistent with the product reference temperature time series information.

[0095] First, when the product temperature prediction time series information output by the temperature prediction unit is inconsistent with the product reference temperature time series information (i.e., the ideal product temperature change curve), it indicates that the current temperature update control parameter cannot achieve the expected temperature control effect. At this time, the parameter iteration unit adjusts the temperature update control parameter and starts a new round of prediction evaluation loop, continuously trying different parameter combinations to find the optimal control parameter. Subsequently, the parameter set statistics unit monitors the progress of the iteration process. When the number of iteration cycles reaches the preset cycle number limit, the parameter set statistics unit terminates the iteration process and collects all the analyzed temperature update control parameters and their corresponding prediction results to form an analyzed temperature control parameter set. This set contains a large amount of control effect information under different parameter combinations, providing a rich data basis for subsequent parameter optimization.

[0096] Next, the dynamic distance evaluation unit evaluates the product temperature prediction timing information corresponding to each set of parameters in the analyzed temperature control parameter set. The dynamic distance evaluation unit uses a dynamic time warping distance algorithm to compare the similarity between each set of product temperature prediction timing information and the product reference temperature timing information. The dynamic time warping algorithm can effectively handle the nonlinear deformation and time axis expansion problems of time series data, and provide a more accurate measure for the similarity evaluation of timing information. Through this evaluation, the dynamic distance evaluation unit obtains a dynamic time warping distance set that characterizes the control effect of each set of parameters. Then, the parameter sorting unit sorts the analyzed temperature control parameter set according to the above dynamic time warping distance set. Specifically, the parameter sorting unit rearranges all the analyzed temperature control parameters in the order of dynamic time warping distance from small to large, and obtains the sequence of analyzed temperature control parameter sets. The smaller the distance, the closer the temperature prediction timing generated by the corresponding parameter is to the reference temperature timing, and the better the control effect.

[0097] Subsequently, the preferred parameter extraction unit extracts the top k temperature control parameters (i.e., the k groups of parameters with the best control effect) from the sorted sequence of analyzed temperature control parameter sets. For these k groups of parameters, the preferred parameter extraction unit classifies the parameters according to different attributes, calculates the mean of the parameters with the same attributes, and obtains the adjustment endpoint temperature control parameters that integrate the characteristics of multiple groups of excellent parameters, thereby realizing the fusion of the characteristics of multiple groups of excellent parameters, which helps to obtain more robust control parameters. Afterwards, the parameter expansion unit uses the above-mentioned adjustment endpoint temperature control parameters as a guide to adjust the bottom 20% of the parameters (i.e., the parameters with poor control effects) in the analyzed temperature control parameter set sequence. The adjustment process brings these parameters closer to the adjustment endpoint temperature control parameters while maintaining a certain degree of randomness to form a new set of expanded temperature control parameters. These expanded parameters will be sent to a new round of cyclic evaluation to achieve continuous exploration and optimization of the parameter space.

[0098] Through the iterative optimization mechanism, the optimal control parameters can be efficiently searched in the complex parameter space to provide the best initialization configuration for the temperature controller.

[0099] Embodiment 2: The embodiment of the present invention also provides an industrial temperature intelligent monitoring method of the Internet of Things, such as Figure 2 As shown, including:

[0100] Receive the time series value of the environmental monitoring temperature and the pre-stored value of the environmental temperature to perform deviation weight distribution to obtain the time series value of the monitoring temperature weight;

[0101] According to the monitoring temperature weighted time series value, weighting the deviation between the environment monitoring temperature time series value and the environment temperature pre-stored value to obtain a temperature deviation coefficient;

[0102] Constrained by the target product model and production control parameters, retrieve the proportion of samples with temperature fluctuations less than the temperature fluctuation threshold among the same-constraint samples that meet the temperature deviation coefficient, and set it as the temperature controllability coefficient;

[0103] When the temperature controllability coefficient is greater than or equal to the temperature controllability coefficient threshold, initialize the temperature controller according to the preset temperature control parameters;

[0104] When the temperature controllability coefficient is less than the temperature controllability coefficient threshold, optimize the preset temperature control parameters based on the environmental monitoring temperature time series values to obtain the target temperature control parameters to initialize the temperature controller, and at the same time update the pre-stored environmental temperature value with the centralized temperature value of the environmental monitoring temperature time series values.

[0105] Furthermore, receive the environmental monitoring temperature time series values and the pre-stored environmental temperature value for deviation weight allocation to obtain the monitored temperature weight time series values, including:

[0106] Perform neighborhood hierarchical clustering on the environmental monitoring temperature time series values to obtain a sequence of environmental temperature clustering values;

[0107] Traverse the sequence of environmental temperature clustering values and calculate the sequence of temperature deviation values from the pre-stored environmental temperature value;

[0108] Traverse the sequence of temperature deviation values and calculate the ratio to the sum of the temperature deviations of the sequence of temperature deviation values to obtain the monitored temperature weight time series values.

[0109] Furthermore, constrained by the target product model and production control parameters, retrieve the proportion of samples with temperature fluctuations less than the temperature fluctuation threshold among the same-constraint samples that meet the temperature deviation coefficient, and set it as the temperature controllability coefficient, including:

[0110] Construct a query constraint condition based on the target product model and the production control parameters;

[0111] Retrieve the pre-stored environmental temperature record values, environmental temperature monitoring record time series values, product temperature pre-stored record time series values, and product temperature monitoring record time series values of the first same-constraint samples that meet the query constraint condition;

[0112] Calculate the sample temperature deviation coefficient of the pre-stored environmental temperature record value and the environmental temperature monitoring record time series value;

[0113] When the deviation between the sample temperature deviation coefficient and the temperature deviation coefficient is less than or equal to the temperature deviation coefficient tolerance threshold, calculate the proportion of time series with temperature deviations greater than or equal to the temperature deviation threshold between the product temperature pre-stored record time series value and the product temperature monitoring record time series value, and set it as the first temperature fluctuation value;

[0114] Until the Qth temperature fluctuation value is obtained, where Q is an integer and Q ≥ 500;

[0115] Statistically analyze the proportion of samples less than the temperature fluctuation threshold among the first temperature fluctuation value to the Qth temperature fluctuation value, and set it as the temperature controllability coefficient.

[0116] Furthermore, the embodiments of the present invention further include:

[0117] When the number of the first co-constraint samples that meet the query constraint conditions is less than 2Q, obtain the sample production control parameters of the first co-constraint samples;

[0118] Based on the target product model and the sample production control parameters, construct a query constraint update condition;

[0119] Retrieve the second co-constraint samples that meet the query constraint update conditions;

[0120] Combining the first co-constraint samples and the second co-constraint samples, calculate the temperature controllability coefficient.

[0121] Furthermore, when the temperature controllability coefficient is less than the temperature controllability coefficient threshold, optimize the preset temperature control parameters based on the environmental monitoring temperature time series values to obtain target temperature control parameters, including:

[0122] Obtain a product temperature predictor bound to the target product model and the temperature controller model;

[0123] Statistically analyze the centralized temperature values of the environmental monitoring temperature time series values;

[0124] Update the preset temperature control parameters to obtain temperature update control parameters;

[0125] Input the centralized temperature values and the temperature update control parameters into the product temperature predictor, and output product temperature prediction time series information;

[0126] When the product temperature prediction time series information is consistent with the product reference temperature time series information, set the temperature update control parameters as the temperature control parameters.

[0127] Furthermore, obtaining a product temperature predictor bound to the target product model and the temperature controller model includes:

[0128] Taking the target product model and the temperature controller model as static constraints, taking the environmental temperature and temperature control parameters as independent variables, and taking the product temperature as the dependent variable, collect multiple groups of data;

[0129] Construct a first loss function, where the first loss function is used to count the proportion of moments when the temperature deviation between the product temperature supervision time series information and the product temperature prediction time series information is greater than or equal to the temperature deviation threshold, and the first loss function has a first weight;

[0130] Construct a second loss function, where the second loss function is used to count the average temperature deviation between the product temperature supervision time series information and the product temperature prediction time series information, and the second loss function has a second weight;

[0131] According to the first weight and the second weight, add the first loss function and the second loss function to obtain a loss function, retrieve the multiple sets of data, and train the product temperature predictor.

[0132] Further, the embodiment of the present invention further includes:

[0133] When the product temperature prediction time series information is inconsistent with the product reference temperature time series information, adjust the temperature update control parameter to execute a loop;

[0134] Until the number of loops meets the preset number of loops, obtain the analyzed temperature control parameter set;

[0135] Traverse the product temperature prediction time series information set of the analyzed temperature control parameter set, and perform dynamic time warping distance evaluation with the product reference temperature time series information to obtain a dynamic time warping distance set;

[0136] Sort the analyzed temperature control parameter set in ascending order according to the dynamic time warping distance set to obtain an analyzed temperature control parameter set sequence;

[0137] Extract the top k analyzed temperature control parameters in the sorted analyzed temperature control parameter set sequence, perform the calculation of the mean value of the same attribute parameters, and obtain the adjusted end temperature control parameter;

[0138] Guided by the adjusted end temperature control parameter, adjust the 20% of the analyzed temperature control parameters at the end of the sorted analyzed temperature control parameter set sequence to obtain an extended temperature control parameter set and execute a loop.

[0139] It should be noted that in the above embodiments, the descriptions of the respective embodiments have their own emphases. For parts not described in detail in a certain embodiment, reference may be made to the relevant descriptions of other embodiments.

[0140] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk memory, CD-ROM, optical memory, etc.) that contain computer-usable program code.

[0141] The present invention is described with reference to the flowcharts and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the present invention. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and combinations of flows and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded computer, or other programmable data processing devices to produce a machine, such that the instructions executed by the processor of the computer or other programmable data processing devices produce means for implementing the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0142] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory produce a manufactured article including instruction means that implement the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0143] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operation steps are performed on the computer or other programmable device to produce a computer-implemented process, so that the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0144] Although the preferred embodiments of the present invention have been described, those skilled in the art can make additional changes and modifications to these embodiments once they know the basic creative concepts.

[0145] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the present invention and its equivalent technologies, the present invention is also intended to include these modifications and variations.

Claims

1. An industrial temperature intelligent monitoring system for the Internet of Things, characterized in that, Including: A deviation weight allocation module, configured to receive the environmental monitoring temperature time series value and the pre - stored environmental temperature value for deviation weight allocation to obtain the monitored temperature weight time series value; A deviation weighting module, configured to weight the deviation between the environmental monitoring temperature time series value and the pre - stored environmental temperature value according to the monitored temperature weight time series value to obtain a temperature deviation coefficient; A controllability evaluation module, configured to retrieve, with the target product model and production control parameters as constraints, the proportion of samples whose temperature fluctuations of the same - constraint samples meeting the temperature deviation coefficient are less than the temperature fluctuation threshold, and set it as the temperature controllability coefficient; A first initialization module, configured to initialize the temperature controller according to the preset temperature control parameters when the temperature controllability coefficient is greater than or equal to the temperature controllability coefficient threshold; A second initialization module, configured to optimize the preset temperature control parameters based on the environmental monitoring temperature time series value when the temperature controllability coefficient is less than the temperature controllability coefficient threshold, obtain the target temperature control parameters to initialize the temperature controller, and update the pre - stored environmental temperature value with the central temperature value of the environmental monitoring temperature time series value; Among them, the deviation weight allocation module includes: A temperature clustering unit, configured to perform neighborhood - level clustering on the environmental monitoring temperature time series value to obtain an environmental temperature clustering value sequence; A deviation calculation unit, configured to traverse the environmental temperature clustering value sequence and calculate a temperature deviation value sequence from the pre - stored environmental temperature value; A weight allocation unit, configured to traverse the temperature deviation value sequence and calculate the ratio to the sum of temperature deviations of the temperature deviation value sequence to obtain the monitored temperature weight time series value; Among them, the controllability evaluation module includes: A constraint condition construction unit, configured to construct a query constraint condition based on the target product model and the production control parameters; A sample retrieval unit, configured to retrieve the pre - stored environmental temperature record value, the environmental temperature monitoring record time series value, the pre - stored product temperature record time series value, and the product temperature monitoring record time series value of the first same - constraint samples meeting the query constraint condition; A sample deviation analysis unit, configured to calculate the sample temperature deviation coefficient of the pre - stored environmental temperature record value and the environmental temperature monitoring record time series value; A temperature fluctuation analysis unit, configured to calculate the proportion of time series with temperature deviations of the pre - stored product temperature record time series value and the product temperature monitoring record time series value greater than or equal to the temperature deviation threshold as the first temperature fluctuation value when the deviation between the sample temperature deviation coefficient and the temperature deviation coefficient is less than or equal to the temperature deviation coefficient tolerance threshold; until the Q - th temperature fluctuation value is obtained, where Q is an integer and Q≥500; A sample statistics unit, configured to statistically calculate the proportion of samples less than the temperature fluctuation threshold among the first temperature fluctuation value to the Q - th temperature fluctuation value, and set it as the temperature controllability coefficient.

2. The system according to claim 1, wherein The controllability evaluation module further includes: A sample supplement trigger unit, configured to obtain the sample production control parameters of the first same - constraint samples when the number of the first same - constraint samples meeting the query constraint condition is less than 2Q; A constraint update unit, configured to construct a query constraint update condition based on the target product model and the sample production control parameters; A supplementary sample retrieval unit for retrieving second same-constraint samples that meet the query constraint update conditions; A comprehensive calculation unit for calculating the temperature controllable coefficient by combining the first same-constraint sample and the second same-constraint sample; 3. The system according to claim 1, characterized in that The second initialization module includes: A predictor acquisition unit for obtaining a product temperature predictor bound to the target product model and the temperature controller model; A centralized temperature statistics unit for statistically calculating the centralized temperature value of the environmental monitoring temperature time series values; A parameter update unit for updating the preset temperature control parameters to obtain temperature update control parameters; A temperature prediction unit for inputting the centralized temperature value and the temperature update control parameters into the product temperature predictor and outputting product temperature prediction time series information; A parameter confirmation unit for setting the temperature update control parameters as the temperature control parameters when the product temperature prediction time series information is consistent with the product reference temperature time series information; 4. The system according to claim 3, characterized in that, The predictor acquisition unit includes: A data acquisition sub-unit for collecting multiple groups of data with the target product model and the temperature controller model as static constraints, environmental temperature and temperature control parameters as independent variables, and product temperature as the dependent variable; A first function construction unit for constructing a first loss function, where the first loss function is used to statistically calculate the proportion of moments when the temperature deviation between the product temperature supervision time series information and the product temperature prediction time series information is greater than or equal to the temperature deviation threshold, and the first loss function has a first weight; A second function construction unit for constructing a second loss function, where the second loss function is used to statistically calculate the mean value of the temperature deviation between the product temperature supervision time series information and the product temperature prediction time series information, and the second loss function has a second weight; A predictor training sub-unit for adding the first loss function and the second loss function according to the first weight and the second weight to obtain a loss function, retrieving the multiple groups of data, and training the product temperature predictor; 5. The system according to claim 3, wherein The second initialization module further includes: A parameter iteration unit for adjusting the temperature update control parameters to execute a loop when the product temperature prediction time series information is inconsistent with the product reference temperature time series information; A parameter set statistics unit for obtaining an analyzed temperature control parameter set until the number of loops meets the preset number of loops; A dynamic distance evaluation unit for traversing the product temperature prediction time series information set of the analyzed temperature control parameter set and performing a dynamic time warping distance evaluation with the product reference temperature time series information to obtain a dynamic time warping distance set; A parameter sorting unit for sorting the analyzed temperature control parameter set in ascending order according to the dynamic time warping distance set to obtain an analyzed temperature control parameter set sequence; An optimal parameter extraction unit for extracting the top k analyzed temperature control parameters in the sorted analyzed temperature control parameter set sequence, performing a mean calculation of the same-attribute parameters, and obtaining adjusted end temperature control parameters; A parameter expansion unit, which is used to adjust 20% of the analyzed temperature control parameters with relatively late sorting in the analyzed temperature control parameter set sequence under the guidance of the adjusted end temperature control parameter, so as to obtain an expanded temperature control parameter set execution loop.

6. An industrial temperature intelligent monitoring method for the Internet of Things, characterized in that, Implemented by the industrial temperature intelligent monitoring system of the Internet of Things according to any one of claims 1 to 5, including: Receiving the environmental monitoring temperature time series value and the environmental temperature pre-stored value for deviation weight allocation to obtain the monitored temperature weight time series value; According to the monitored temperature weight time series value, weighting the deviation between the environmental monitoring temperature time series value and the environmental temperature pre-stored value to obtain a temperature deviation coefficient; Constrained by the target product model and production control parameters, retrieving the proportion of samples with temperature fluctuations less than the temperature fluctuation threshold among the samples with the same constraints that meet the temperature deviation coefficient, and setting it as the temperature controllable coefficient; When the temperature controllable coefficient is greater than or equal to the temperature controllable coefficient threshold, initializing the temperature controller according to the preset temperature control parameter; When the temperature controllable coefficient is less than the temperature controllable coefficient threshold, optimizing the preset temperature control parameter based on the environmental monitoring temperature time series value to obtain the target temperature control parameter to initialize the temperature controller, and at the same time updating the environmental temperature pre-stored value with the centralized temperature value of the environmental monitoring temperature time series value.

Images