Temperature controller nonlinear compensation control method and system based on fuzzy logic

Abstract

The invention relates to the technical field of temperature control, in particular to a temperature controller nonlinear compensation control method and system based on fuzzy logic, and the method comprises the steps: constructing a double-closed-loop control architecture of a temperature control system; main control data is generated through the outer loop fuzzy decision layer, and feedforward compensation and lag compensation processing is carried out on the main control data through the inner loop compensation execution layer to obtain a final control quantity; and adjusting operation parameters of the temperature control system according to the final control quantity. Through a double compensation mechanism of feed-forward compensation and lag compensation, accurate pre-judgment and self-adaptive adjustment of control data are realized, and the problems of control lag and nonlinear instability caused by sensor delay, inertia of an execution mechanism and environmental interference in a temperature control system are systematically solved.

Description

Technical Field

[0001] This application relates to the technical field of temperature control, and particularly to a non-linear compensation control method and system for a thermostat based on fuzzy logic. Background Art

[0002] With the rapid development of industrial automation and intelligence, thermostats are increasingly widely used in various systems. However, traditional thermostats often have problems of insufficient non-linear compensation under complex working conditions, resulting in problems such as decreased control accuracy, slower response speed, and increased energy consumption. Especially in a dynamic environment change, the performance of the thermostat will be significantly affected, making it difficult to meet the requirements of high-precision and high-efficiency modern industries.

[0003] After retrieval, the Chinese invention patent with the publication number CN110426950B discloses an intelligent robot based on fuzzy logic. This patent adjusts the parameters of the neural network module through a fuzzy logic module to achieve precise control of the speed of the robot's driving wheels. However, this technical solution mainly focuses on robot motion control. The design of its fuzzy logic module does not fully consider the problem of compensating for the unique non-linear characteristics of the thermostat, and lacks an optimization mechanism for the dynamic response to temperature changes. In addition, the support for real-time adaptability in a complex environment of this solution is limited, which may lead to the inability to effectively cope with rapidly changing temperature conditions in a thermostat scenario.

[0004] After retrieval, the Chinese invention patent with the publication number CN103823368B discloses a PID-type fuzzy logic control method based on a weight rule table. This patent simplifies the implementation process of fuzzy logic control through a weight rule table and optimizes the overshoot and oscillation problems of the control system. However, this technical solution is not specifically designed for the non-linear compensation requirements of the thermostat. The setting of its weight rule table depends on expert experience, and may lack sufficient flexibility and adaptability when facing complex thermostat scenarios. In addition, this solution does not fully consider the impact of temperature sensor data delay on control accuracy, which may lead to large control deviations in actual applications.

[0005] The above problems indicate that the existing fuzzy logic-based control technologies have deficiencies such as insufficient non-linear compensation, lack of optimization of dynamic response, and limited adaptability to complex environments when applied to thermostats. Therefore, the present invention provides a non-linear compensation control method and system for a thermostat based on fuzzy logic, aiming to improve the control accuracy, response speed, and energy consumption efficiency of the thermostat under complex working conditions by introducing a fuzzy logic optimization mechanism for the characteristics of the thermostat, so as to meet the requirements of modern industries for efficient and intelligent thermostats. Summary of the Invention

[0006] To solve the above problems, the present invention provides a non-linear compensation control method and system for a thermostat based on fuzzy logic. Through a dual compensation mechanism of feed-forward compensation and lag compensation, precise prediction and adaptive adjustment of control data are achieved, systematically solving the problems of control lag and non-linear instability in the temperature control system caused by sensor delay, actuator inertia, and environmental interference.

[0007] The object of the present invention can be achieved through the following technical solutions: In the first aspect, the present invention provides a non-linear compensation control method for a thermostat based on fuzzy logic, including the steps of: S1. Construct a double-closed-loop control architecture for the temperature control system, where the double-closed-loop control architecture includes an outer-loop fuzzy decision-making layer and an inner-loop compensation execution layer; S2. Generate main control data through the outer-loop fuzzy decision-making layer, and perform feed-forward compensation and lag compensation processing on the main control data through the inner-loop compensation execution layer to obtain the final control quantity; S3. Adjust the operating parameters of the temperature control system according to the final control quantity; The generation of the main control data through the outer-loop fuzzy decision-making layer includes the steps of: Collect the temperature and humidity deviation values, and adjust the weights of the temperature main control rule set and the humidity main control rule set; Perform the first normalization process on the temperature and humidity deviation values based on the adjusted weights of the temperature main control rule set and the humidity main control rule set, and use the obtained data as input variables; Map the input variables to fuzzy language variables according to the membership function as the main control data.

[0008] As a preferred technical solution of the present invention, the adjustment of the weights of the temperature main control rule set and the humidity main control rule set includes the steps of: Adjust the weights of the temperature main control rule set and the humidity main control rule set according to the temperature and humidity coupling coefficient. Specifically, when the temperature and humidity coupling coefficient is greater than the preset coupling threshold, increase the weight of the humidity main control rule set and decrease the weight of the temperature main control rule set; when the temperature and humidity coupling coefficient is less than or equal to the preset coupling threshold, maintain the original weights; Among them, the calculation formula of the temperature and humidity coupling coefficient is: ; The calculation formula of the weight of the temperature main control rule set is: ; The calculation formula of the weight of the humidity main control rule set is: ; In the formula, is the temperature deviation and the humidity deviation The covariance of is the standard deviation of the temperature deviation, is the standard deviation of the humidity deviation, is the temperature-humidity coupling coefficient, is the weight of the temperature main control rule set before adjustment, is the weight of the temperature main control rule set after adjustment, is the weight of the humidity main control rule set after adjustment.

[0009] As a preferred technical solution of the present invention, the first normalization process includes the steps of: According to the dynamic characteristics of the temperature control system, several fuzzy sets are established, and the input variables are divided into the fuzzy sets; the fuzzy sets are used to refine the degree of temperature-humidity deviation; The formula for the first normalization process is: ; wherein, is the temperature deviation, representing the deviation between the current temperature and the set temperature, is the minimum value of the temperature deviation, is the maximum value of the temperature deviation, is the humidity deviation, representing the deviation between the current humidity and the set humidity, is the minimum value of the humidity deviation, is the maximum value of the humidity deviation, is the value obtained by the first normalization process of the temperature-humidity deviation value.

[0010] As a preferred technical solution of the present invention, the process of performing feedforward compensation and lag compensation on the main control data through the inner loop compensation execution layer includes the steps of: Extract the heat source characteristics strongly correlated with the temperature-humidity deviation value; the heat source characteristics include the equipment start-stop state, the environmental temperature change trend, and the historical interference data; Taking the error between the historical main control data and the actual execution result as the supervision signal, training the LSTM model according to the supervision signal, and generating the compensation amount of the actual execution result for the ideal historical main control data result; Input the extracted heat source characteristics into the LSTM model according to the time series, and generate the interference suppression amount for the inertia of the actuator and the sensor delay; use the interference suppression amount as the pre-compensation amount for the main control data; map the interference suppression amount to a fuzzy linguistic variable according to the interference membership function, and generate interference suppression data; Superimpose the interference suppression data and the main control data to generate intermediate control data; Calculate the time lag to be compensated, and dynamically correct the intermediate control data according to the time lag to be compensated to generate the final control data.

[0011] As a preferred technical solution of the present invention, mapping the interference suppression amount to a fuzzy linguistic variable according to the interference membership function to generate interference suppression data includes the steps of: Map the interference suppression amount to the interference fuzzy set; Map the interference suppression amount to the interval [0, 1] through the second normalization process; Generate interference suppression data from the interference suppression amount through the interference membership function.

[0012] As a preferred technical solution of the present invention, superimposing the interference suppression data and the main control data to generate intermediate control data includes the steps of: Obtain the main control data generated by the outer loop fuzzy decision layer; Unify the dimensions of the main control data and the interference suppression data; Generate intermediate control data using the linear superposition formula; The linear superposition formula is: ; Where, is the main control data; is the interference suppression data; is the superposition coefficient; is the intermediate control data.

[0013] As a preferred technical solution of the present invention, calculating the time delay to be compensated and dynamically correcting the intermediate control data according to the time delay to be compensated includes the steps of: Extract the response sequences of the main control data and the intermediate control data; Calculate the phase lag angle between the main control data and the system response; Calculate the pure time delay according to the phase lag angle, and at the same time, inversely deduce the actual time delay through the time sequence offset of the main control data and the intermediate control data; Subtract the pure time delay from the actual time delay to obtain the time delay to be compensated; Combine the intermediate control data and the time delay to be compensated to generate the final control data; the formula for generating the final control data is: ; Where, is the time delay to be compensated, is the final control data, is the intermediate control data applied in advance

[0014] Second aspect, the present invention also provides a non-linear compensation control system for a thermostat based on fuzzy logic, which executes the non-linear compensation control method for a thermostat based on fuzzy logic as described above. The system includes a data acquisition module, an outer loop decision-making module, an inner loop compensation module, and a strategy optimization module; The data acquisition module is used to obtain the temperature and humidity deviation values in real time and construct a multi-dimensional input data set; The outer loop decision-making module executes fuzzy reasoning and decision generation based on a double closed-loop control architecture, including a weight adjustment unit and a first normalization processing unit; the weight adjustment unit is used to dynamically adjust the weights of the temperature main control rule set and the humidity main control rule set to solve the coupling interference during the joint adjustment of temperature and humidity; the first normalization processing unit is used to eliminate the dimensional difference of the temperature and humidity deviation values through a non-linear mapping algorithm; The inner loop compensation module is used to implement a double compensation mechanism for the inertia of the actuator and the delay of the sensor; it includes a feed-forward compensation unit and a lag compensation unit; The strategy optimization module is used to perform temperature control compensation according to the interference suppression amount and the lag time to be compensated.

[0015] As a preferred technical solution of the present invention, the feed-forward compensation unit learns the error between the historical main control data and the actual execution effect through an LSTM model, and generates an interference suppression amount for the inertia of the actuator and the delay of the sensor.

[0016] As a preferred technical solution of the present invention, the lag compensation unit generates the lag time to be compensated according to the phase lag angle calculation and the pure lag time back-calculation technology, and realizes lag compensation through time axis translation.

[0017] The present invention provides a non-linear compensation control method and system for a thermostat based on fuzzy logic, and has the following beneficial effects: Through the double compensation mechanism of feed-forward compensation and lag compensation, the present invention realizes the accurate prediction and adaptive adjustment of control data, and systematically solves the problems of control lag and non-linear instability caused by sensor delay, actuator inertia and environmental interference in the temperature control system; by learning the error between the historical main control data and the actual execution effect through an LSTM model, an interference suppression amount for the inertia of the actuator and the delay of the sensor is generated to offset the control lag effect caused by physical delay; according to the phase lag angle calculation and the pure lag time back-calculation technology, the lag time to be compensated is generated, and lag compensation is realized through time axis translation, so as to offset the total lag effect of the system; effectively solves the limitation that traditional fixed lag compensation cannot adapt to time-varying working conditions, and significantly improves the robustness and steady-state accuracy of the control system. Description of the Drawings

[0018] To more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the accompanying drawings required in the embodiments or the description of the prior art. Obviously, the accompanying drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other accompanying drawings can also be obtained based on these drawings.

[0019] Figure 1 It is the first process schematic diagram of the non - linear compensation control method for a thermostat based on fuzzy logic provided by the embodiments of the present invention; Figure 2 It is the second process schematic diagram of the non - linear compensation control method for a thermostat based on fuzzy logic provided by the embodiments of the present invention; Figure 3 It is the structural schematic diagram of the non - linear compensation control system for a thermostat based on fuzzy logic provided by the embodiments of the present invention. Detailed implementation manners

[0020] The following will clearly and completely describe the technical solutions in the embodiments of the present application with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, rather than all embodiments. Based on the embodiments of the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts also belong to the scope protected by the present application.

[0021] The following further describes the optimal embodiments of the present invention with reference to the accompanying drawings; Embodiment 1 Please refer to Figure 1 , this embodiment provides a non - linear compensation control method for a thermostat based on fuzzy logic, including: S1. Construct a double - closed - loop control architecture for the temperature control system, and the double - closed - loop control architecture includes an outer - loop fuzzy decision - making layer and an inner - loop compensation execution layer; S2. Generate main control data through the outer - loop fuzzy decision - making layer, and perform feed - forward compensation and lag compensation processing on the main control data through the inner - loop compensation execution layer to obtain the final control quantity; S3. Adjust the operating parameters of the temperature control system according to the final control quantity.

[0022] It is understandable that the double closed-loop control architecture consists of an outer-loop fuzzy decision-making layer and an inner-loop compensation execution layer. The outer-loop fuzzy decision-making layer is responsible for generating main control data, and the inner-loop compensation execution layer is responsible for performing feed-forward compensation and lag compensation processing on the main control data to obtain the final control quantity. The outer-loop fuzzy decision-making layer and the inner-loop compensation execution layer achieve collaborative operation through information flow transmission. Among them, the outer-loop fuzzy decision-making layer transmits the main control data to the inner-loop compensation execution layer, and the inner-loop compensation execution layer performs feed-forward compensation and lag compensation processing according to the main control data to generate the final control quantity. The system adjusts the operating parameters of the temperature control system according to the final control quantity and feeds back the real-time monitoring data to the outer-loop fuzzy decision-making layer to update the fuzzy rule base and membership function parameters.

[0023] Specifically, generating the main control data through the outer-loop fuzzy decision-making layer includes the steps of: The fuzzy rule set includes a temperature main control rule set and a humidity main control rule set; Collect the temperature and humidity deviation values and adjust the weights of the temperature main control rule set and the humidity main control rule set; Perform the first normalization processing on the temperature and humidity deviation values based on the adjusted weights of the temperature main control rule set and the humidity main control rule set, and use the obtained data as the input variables; Map the input variables to fuzzy linguistic variables according to the membership function as the main control data.

[0024] Based on the foregoing, adjusting the weights of the temperature main control rule set and the humidity main control rule set to solve the coupling interference during temperature and humidity joint adjustment includes the steps of: The weights of the temperature main control rule set and the humidity main control rule set are adjusted according to the temperature and humidity coupling coefficient; when the temperature and humidity coupling coefficient is greater than the preset coupling threshold, increase the weight of the humidity main control rule set and decrease the weight of the temperature main control rule set; when the temperature and humidity coupling coefficient is less than or equal to the preset coupling threshold, maintain the original weight; The calculation formula for the temperature and humidity coupling coefficient is: ; The calculation formula for the weight of the temperature main control rule set is: ; The calculation formula for the weight of the humidity main control rule set is: ; Among them, is the covariance of the temperature deviation and the humidity deviation, is the standard deviation of the temperature deviation, is the standard deviation of the humidity deviation, is the temperature and humidity coupling coefficient, is the weight of the temperature main control rule set before adjustment, is the weight of the temperature main control rule set after adjustment, is the weight of the humidity main control rule set after adjustment.

[0025] Based on the foregoing, the first normalization process is used to eliminate the dimensional difference of the temperature and humidity deviation values and map them uniformly to a specified interval, including: According to the dynamic characteristics of the temperature control system, several fuzzy sets are established and the input variables are divided into the fuzzy sets; The membership function is used to convert the input variables into fuzzy linguistic variables, providing a basis for fuzzy inference; the fuzzy sets are used to refine the degree of temperature and humidity deviation, so as to map the values obtained by the first normalization of the temperature and humidity deviation values into each fuzzy set; The core formula of the first normalization process is: ; where, is the deviation between the current temperature and the set temperature, is the minimum possible value of the temperature deviation, is the maximum possible value of the temperature deviation, is the deviation between the current humidity and the set humidity, is the minimum possible value of the humidity deviation, is the maximum possible value of the humidity deviation, is the value obtained by the first normalization of the temperature and humidity deviation values.

[0026] Based on the foregoing, in this embodiment, the data obtained by the first normalization process is mapped to the interval [-1, 1], and 5 fuzzy sets are established, namely negative large, negative small, zero, positive small, and positive large; when the input variable is in the range [-1, -0.6), the fuzzy set is negative large; when the input variable is in the range [-0.6, -0.2), the fuzzy set is negative small; when the input variable is in the range [-0.2, 0.2), the fuzzy set is zero; when the input variable is in the range [0.2, 0.6), the fuzzy set is positive small; when the input variable is in the range [0.6, 1], the fuzzy set is positive large; Then, the membership function is used to calculate the membership degrees of the input variables to each fuzzy set, realizing the fuzzification of the input variables; When the input variable is in the range [-1, -0.6), the membership degree is ; When the input variable is in the range [-0.6, -0.2), the membership degree is ; When the input variable is in the range [-0.2, 0.2), the membership degree is ; When the input variable is in the range [0.2, 0.6), the membership degree is ; When the input variable is in the range of [0.6, 1], the membership degree is ; Among them, , , and are all function parameters used to define the boundaries and slopes of the fuzzy sets, and x is the input variable.

[0027] At this time, the main control data is It can be understood that in this embodiment, by adjusting the weights of the temperature main control rule set and the humidity main control rule set, the problem that the traditional single-layer rule base cannot distinguish the primary and secondary interferences of temperature and humidity is avoided, and the robustness of the system to complex working conditions is further enhanced; by the first normalization process, the dimensional differences of the temperature and humidity deviation values are eliminated, so that they can participate in the operation on the same scale and the control deviation caused by dimensional differences is avoided; by fuzzyfying the input variables, it is convenient for subsequent logical reasoning.

[0028] Specifically, the process of performing feedforward compensation and lag compensation on the main control data by the inner-loop compensation execution layer includes the steps of: Extracting heat source features strongly related to the temperature and humidity deviation values; the heat source features include the equipment start-stop state, the environmental temperature change trend, and historical interference data; Using the error between the historical main control data and the actual execution result as the supervision signal, training the LSTM model according to the supervision signal, and generating the compensation amount of the actual execution result for the ideal historical main control data result; Inputting the extracted heat source features into the LSTM model according to the time series to generate an interference suppression amount for the inertia of the actuator and the sensor delay; using the interference suppression amount as the pre-compensation amount for the main control data; mapping the interference suppression amount to a fuzzy linguistic variable according to the interference membership function to generate interference suppression data; Superimposing the interference suppression data and the main control data to generate intermediate control data; Calculating the lag time to be compensated, and dynamically correcting the intermediate control data according to the lag time to be compensated to generate the final control data.

[0029] Based on the foregoing, mapping the interference suppression amount to a fuzzy linguistic variable according to the interference membership function to generate interference suppression data, including the steps of: Mapping the interference suppression amount to the interference fuzzy set: the interference fuzzy set includes no disturbance, low compensation, medium compensation, and high compensation; when there is no disturbance, no compensation is required; when there is low compensation, the compensation amount is small and is suitable for small deviation situations; when there is medium compensation, the compensation amount is moderate and is suitable for medium deviations; when there is high compensation, the compensation amount is large and is suitable for large deviations or emergency situations; Mapping the interference suppression amount to the interval [0, 1] through the second normalization process, for example: When the interference suppression amount is in the range of [0, 0.4), the fuzzy set is no perturbation; the superposition coefficient , and the interference membership function is 0; When the interference suppression amount is in the range of [0.4, 0.6), the fuzzy set is low compensation, and the superposition coefficient , and the interference membership function is ; When the interference suppression amount is in the range of [0.6, 0.8), the fuzzy set is medium compensation, and the superposition coefficient , and the interference membership function is ; When the interference suppression amount is in the range of [0.8, 1], the fuzzy set is high compensation, and the superposition coefficient , and the interference membership function is ; Through the interference membership function, the interference suppression amount is used to generate interference suppression data.

[0030] Based on the foregoing, the process of superimposing the interference suppression data and the main control data to generate intermediate control data includes the following steps: Obtain the main control data generated by the outer loop fuzzy decision layer; Unify the dimensions of the main control data and the interference suppression data; Use a linear superposition formula to generate intermediate control data; The linear superposition formula is: ; Wherein, is the main control data; is the interference suppression data; is the superposition coefficient; is the intermediate control data.

[0031] Based on the foregoing, the process of calculating the time lag to be compensated and dynamically correcting the intermediate control data according to the time lag to be compensated includes the following steps: Extract the response sequences of the main control data and the intermediate control data; As , , , , where is the control period; Calculate the phase lag angle between the main control data and the system response. The phase lag angle is used to quantify the time delay caused by system inertia, delay, or energy transfer characteristics. For example, when the phase lag angle is 0°, the input and output are completely synchronized and there is no time delay; when the phase lag angle is 90°, the output lags behind the input by one-quarter of the control period; Calculate the pure dead time according to the phase lag angle, and at the same time, inversely deduce the actual dead time through the timing offset between the main control data and the intermediate control data, that is, the time difference between the response sequences of the main control data and the intermediate control data; The pure dead time refers to the fixed time delay of the output signal relative to the input signal in the time dimension; the actual dead time includes the pure dead time and the time delay caused by equipment start-stop, sudden load change, etc.; Subtract the pure dead time from the actual dead time to obtain the dead time to be compensated; Combine the intermediate control data and the dead time to be compensated to generate the final control data; the formula for generating the final control data is: ; Wherein, is the dead time to be compensated, is the final control data, is the intermediate control data applied in advance time.

[0032] It can be understood that in this embodiment, through the dual compensation mechanism of feedforward compensation and lag compensation, the accurate prediction and adaptive adjustment of control data are realized, and the problems of control lag and nonlinear instability caused by sensor delay, actuator inertia and environmental interference in the temperature control system are systematically solved; the LSTM model is used to learn the error between the historical main control data and the actual execution effect, generate the interference suppression amount for actuator inertia and sensor delay, and offset the control lag effect caused by physical delay; according to the phase lag angle calculation and pure dead time inverse deduction technology, generate the dead time to be compensated, and realize lag compensation through time axis translation, so as to offset the total system lag effect; effectively solve the limitation that traditional fixed lag compensation cannot adapt to time-varying working conditions, and significantly improve the robustness and steady-state accuracy of the control system.

[0033] Embodiment 2 Please refer to Figure 3 , this embodiment provides a non-linear compensation control system for a thermostat based on fuzzy logic, which executes the non-linear compensation control method for a thermostat based on fuzzy logic described above. The system includes a data acquisition module, an outer loop decision module, an inner loop compensation module and a strategy optimization module; The data acquisition module is used to obtain the temperature and humidity deviation values in real time and construct a multi-dimensional input data set; the data acquisition module includes a sensor data acquisition unit and a feature extraction unit. Among them, the feature extraction unit separates the primary and secondary interference features of temperature and humidity through a decoupling algorithm and generates standardized input variables; The outer loop decision-making module performs fuzzy reasoning and decision-making generation based on a dual closed-loop control architecture, including a weight adjustment unit and a first normalization processing unit; the weight adjustment unit is used to dynamically adjust the weights of the temperature main control rule set and the humidity main control rule set to solve the coupling interference during the joint adjustment of temperature and humidity; the first normalization processing unit is used to eliminate the dimensional difference of the temperature and humidity deviation values through a non-linear mapping algorithm. The inner loop compensation module implements a dual compensation mechanism for the inertia of the actuator and the sensor delay, including a feed-forward compensation unit and a lag compensation unit; the feed-forward compensation unit learns the error between the historical main control data and the actual execution effect through an LSTM model, generates an interference suppression amount for the inertia of the actuator and the sensor delay, and offsets the control lag effect caused by the physical delay; the lag compensation unit generates the lag time to be compensated according to the phase lag angle calculation and the pure lag time backtracking technology, and realizes the lag compensation through the time axis translation, so as to offset the total lag effect of the system. The strategy optimization module is used to perform temperature control compensation according to the interference suppression amount and the lag time to be compensated.

[0034] The above are only the preferred embodiments of the present application, and are not intended to limit the present application. Any modifications, equivalent replacements, and improvements made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A non-linear compensation control method for a thermostat based on fuzzy logic, characterized in that: Including the following steps: S1. Construct a dual - closed - loop control architecture for the temperature control system. The dual - closed - loop control architecture includes an outer - loop fuzzy decision - making layer and an inner - loop compensation execution layer; S2. Generate main control data through the outer - loop fuzzy decision - making layer, and perform feed - forward compensation and lag compensation processing on the main control data through the inner - loop compensation execution layer to obtain the final control quantity; S3. Adjust the operating parameters of the temperature control system according to the final control quantity; The generating of the main control data through the outer - loop fuzzy decision - making layer includes the steps of: Collect the temperature - humidity deviation value and adjust the weights of the temperature main control rule set and the humidity main control rule set; Perform the first normalization process on the temperature - humidity deviation value based on the adjusted weights of the temperature main control rule set and the humidity main control rule set, and use the obtained data as the input variable; Map the input variable to a fuzzy linguistic variable according to the membership function as the main control data.

2. The non-linear compensation control method of the thermostat based on fuzzy logic according to claim 1, characterized in that: The adjusting of the weights of the temperature main control rule set and the humidity main control rule set includes the steps of: Adjust the weights of the temperature main control rule set and the humidity main control rule set according to the temperature - humidity coupling coefficient. Specifically, when the temperature - humidity coupling coefficient is greater than the preset coupling threshold, increase the weight of the humidity main control rule set and decrease the weight of the temperature main control rule set; when the temperature - humidity coupling coefficient is less than or equal to the preset coupling threshold, maintain the original weights; Among them, the calculation formula of the temperature - humidity coupling coefficient is: ； The calculation formula of the weight of the temperature main control rule set is: ； The calculation formula of the weight of the humidity main control rule set is: ； In the formula, is the temperature deviation and the humidity deviation of the covariance, is the standard deviation of the temperature deviation, is the standard deviation of the humidity deviation, is the temperature-humidity coupling coefficient, is the weight of the temperature main control rule set before adjustment, is the weight of the temperature main control rule set after adjustment, is the weight of the humidity main control rule set after adjustment.

3. The non-linear compensation control method of the thermostat based on fuzzy logic according to claim 1, characterized in that: The first normalization process includes the steps of: According to the dynamic characteristics of the temperature control system, set up several fuzzy sets and divide the input variable into the fuzzy sets; the fuzzy sets are used to refine the degree of temperature - humidity deviation; The formula of the first normalization process is: ； Among them, is the temperature deviation, representing the deviation between the current temperature and the set temperature, is the minimum value of the temperature deviation, is the maximum value of the temperature deviation, is the humidity deviation, representing the deviation between the current humidity and the set humidity, is the minimum value of the humidity deviation, is the maximum value of the humidity deviation, is the value obtained by the first normalization processing of the temperature and humidity deviation value.

4. The non-linear compensation control method of the thermostat based on fuzzy logic according to claim 1, characterized in that: The performing of the feed - forward compensation and lag compensation processing on the main control data through the inner - loop compensation execution layer includes the steps of: Extract the heat - source features strongly related to the temperature - humidity deviation value; the heat - source features include the equipment start - stop state, the environmental temperature change trend, and the historical interference data; Use the error between the historical main control data and the actual execution result as the supervision signal, and train the LSTM model according to the supervision signal to generate the compensation amount of the actual execution result for the ideal historical main control data result; Input the extracted heat - source features into the LSTM model according to the time series to generate the interference suppression amount for the inertia of the actuator and the sensor delay; use the interference suppression amount as the pre - compensation amount for the main control data; map the interference suppression amount to a fuzzy linguistic variable according to the interference membership function to generate interference suppression data; Superimpose the interference suppression data and the main control data to generate intermediate control data; Calculate the time of lag to be compensated, and dynamically correct the intermediate control data according to the time of lag to be compensated to generate the final control data.

5. The non-linear compensation control method of the thermostat based on fuzzy logic according to claim 4, characterized in that: The mapping of the interference suppression amount to a fuzzy linguistic variable according to the interference membership function to generate interference suppression data includes the steps of: Map the interference suppression amount to the interference fuzzy set; Map the interference suppression amount to the interval [0, 1] through the second normalization process; Generate interference suppression data from the interference suppression amount through the interference membership function.

6. The non-linear compensation control method of the thermostat based on fuzzy logic according to claim 5, characterized in that: Overlaying the interference suppression data with the main control data to generate intermediate control data includes the steps of: Obtaining the main control data generated by the outer loop fuzzy decision layer; Unifying the dimensions of the main control data and the interference suppression data; Using a linear superposition formula to generate intermediate control data; The linear superposition formula is: ； Among them, is the main control data; is the interference suppression data; is the superposition coefficient; is the intermediate control data.

7. The non-linear compensation control method of the thermostat based on fuzzy logic according to claim 6, characterized in that: Calculating the lag time to be compensated and dynamically correcting the intermediate control data according to the lag time to be compensated, including the steps of: Extracting the response sequences of the main control data and the intermediate control data; Calculating the phase lag angle between the main control data and the system response; Calculating the pure lag time based on the phase lag angle, and inversely inferring the actual lag time through the time series offset of the main control data and the intermediate control data; Taking the difference between the actual lag time and the pure lag time to obtain the lag time to be compensated; Combining the intermediate control data and the lag time to be compensated to generate the final control data; the formula for generating the final control data is: ； Among them, is the lag time to be compensated, is the final control data, is the intermediate control data for advance time application.

8. A non-linear compensation control system for a thermostat based on fuzzy logic, which executes the non-linear compensation control method for a thermostat based on fuzzy logic as described in claim 7, characterized in that: The system includes a data acquisition module, an outer loop decision module, an inner loop compensation module, and a strategy optimization module; The data acquisition module is used to obtain the temperature and humidity deviation values in real time and construct a multi-dimensional input data set; The outer loop decision module performs fuzzy inference and decision generation based on a double closed-loop control architecture, including a weight adjustment unit and a first normalization processing unit; the weight adjustment unit is used to dynamically adjust the weights of the temperature main control rule set and the humidity main control rule set; the first normalization processing unit is used to eliminate the dimensional difference of the temperature and humidity deviation values through a non-linear mapping algorithm; The inner loop compensation module is used to implement a double compensation mechanism for the inertia of the actuator and the sensor delay; it includes a feed-forward compensation unit and a lag compensation unit; The strategy optimization module is used to perform temperature control compensation according to the interference suppression amount and the lag time to be compensated.

9. The non-linear compensation control system of the thermostat based on fuzzy logic according to claim 8, characterized in that: The feed-forward compensation unit learns the error between the historical main control data and the actual execution effect through an LSTM model to generate an interference suppression amount for the inertia of the actuator and the sensor delay.

10. The non-linear compensation control system of the thermostat based on fuzzy logic according to claim 8, characterized in that: The lag compensation unit generates the lag time to be compensated according to the phase lag angle calculation and the pure lag time inverse inference technique, and realizes lag compensation through time axis translation.

Images