Electric appliance intelligent temperature control method and system based on data analysis

Abstract

The invention discloses an electric appliance intelligent temperature control method and system based on data analysis, and relates to the technical field of intelligent temperature control, and the method comprises the steps: collecting environment data, scanning space thermal field distribution, calculating an initial thermal field uniformity index, and integrating the environment data into a local data set through an edge controller; training a local thermodynamic model, generating an initial symbol equation, calculating a fitting error, and obtaining an equation coefficient and a performance index; carrying out federal symbol equation aggregation at the cloud, outputting a global optimization equation, issuing the global optimization equation to an edge controller, replacing a local thermodynamic model, and adjusting a finite element grid heat conductivity coefficient; a digital twinborn environment is constructed at an edge controller, real-time thermal field regulation and control are carried out, and local symbol regression retraining is triggered according to collection feedback. According to the method, the local thermodynamic model is constructed through data analysis, high-precision and self-adaptive intelligent temperature control is realized by combining federal symbol regression optimization and a digital twinning environment, and the thermal field uniformity and the temperature control energy efficiency are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent temperature control, and particularly to an intelligent temperature control method and system for electrical appliances based on data analysis. Background Art

[0002] With the development of intelligent control and Internet of Things technologies, the temperature control mode of electrical appliances is evolving from traditional mechanical or fixed-rule control to data-driven intelligent regulation. Existing intelligent temperature control technologies mainly rely on the feedback of environmental temperature sensors and adjust by combining preset control logics or simple algorithms. However, these methods usually cannot accurately perceive the spatial thermal field distribution, the temperature control accuracy is limited, and they lack the adaptive ability to complex environmental changes. In addition, most of the existing controls are based on static models and fail to dynamically optimize the temperature control strategy by combining thermodynamics modeling and machine learning technologies. The data acquisition and processing capabilities of the existing technologies are limited, making it difficult to accurately depict the spatial thermal field distribution, resulting in uneven local temperatures; the adjustment of temperature control parameters depends on fixed rules, lacking optimization mechanisms of symbolic regression and federated learning, and the adaptability is insufficient; the cloud collaborative optimization ability is weak, and it fails to efficiently utilize distributed computing for real-time regulation. Summary of the Invention

[0003] In view of the above existing problems, the present invention is proposed.

[0004] Therefore, the present invention provides an intelligent temperature control method for electrical appliances based on data analysis to solve the problem of uneven local temperatures caused by the difficulty in accurately depicting the spatial thermal field distribution.

[0005] To solve the above technical problems, the present invention provides the following technical solutions: In a first aspect, the present invention provides an intelligent temperature control method for electrical appliances based on data analysis, which includes collecting environmental data, scanning the spatial thermal field distribution, calculating the initial thermal field uniformity index, and integrating it into a local data set through an edge controller; training a local thermodynamics model, generating an initial symbolic equation and calculating the fitting error, obtaining the equation coefficients and performance indicators, performing data augmentation, and uploading it to the cloud; performing federated symbolic equation aggregation in the cloud, outputting a globally optimized equation, sending it to the edge controller, replacing the local thermodynamics model, and adjusting the thermal conductivity of the finite element mesh; constructing a digital twin environment in the edge controller, performing real-time thermal field regulation, triggering local symbolic regression retraining according to the collected feedback, and uploading the retrained coefficients to the cloud to participate in the next round of federated optimization.

[0006] As a preferred solution of the intelligent temperature control method for electrical appliances based on data analysis of the present invention, wherein: the federated symbolic equation aggregation refers to selecting a non-dominated solution set through the Pareto front screening algorithm, performing symbolic-level crossover operations and applying Gaussian coefficient perturbations, and performing federated iterative evolution of the equation population.

[0007] As a preferred solution of the intelligent temperature control method for electrical appliances based on data analysis according to the present invention, wherein: adjusting the thermal conductivity of the finite element mesh means dynamically adjusting the thermal conductivity of each mesh according to the change trend of the thermal field uniformity output by the global optimization equation, where the adjustment range of the thermal conductivity is proportional to the sensitivity of the thermal field uniformity to the local temperature, and the maximum adjustment range is controlled by a preset sensitivity coefficient.

[0008] As a preferred solution of the intelligent temperature control method for electrical appliances based on data analysis according to the present invention, wherein: training the local thermodynamic model means performing data cleaning and feature normalization on the local data set, constructing a local thermodynamic model using the polynomial regression algorithm, and training, validating, and optimizing it through the local data set.

[0009] As a preferred solution of the intelligent temperature control method for electrical appliances based on data analysis according to the present invention, wherein: the local symbolic regression retraining means that when the deviation between the actual value of the thermal field uniformity calculated by real-time acquisition and the predicted value of the local thermodynamic model exceeds the retraining threshold, adding Gaussian noise to the historical environmental data to generate an enhanced training set, re-running the symbolic regression algorithm, and obtaining the equation coefficients after retraining.

[0010] As a preferred solution of the intelligent temperature control method for electrical appliances based on data analysis according to the present invention, wherein: constructing a digital twin environment for real-time thermal field regulation, the specific steps are as follows Defining the state space, action space, and reward function, and training the policy network using the proximal policy optimization algorithm; After the training converges, scan the spatial thermal field distribution in real time and select the optimal action according to the current state; Drive the stepper motor to adjust the air outlet angle and the frequency converter to adjust the wind speed, and send a pulse signal to the piezoelectric material controller to change the local thermal conductivity.

[0011] As a preferred solution of the intelligent temperature control method for electrical appliances based on data analysis according to the present invention, wherein: the symbolic-level crossover operation means selecting two parent equations from the non-dominated solution set, exchanging the sub-expressions containing the temperature difference term and the angular trigonometric function term to generate a new equation, and applying a random perturbation to the coefficients of the new equation.

[0012] In a second aspect, the present invention provides an intelligent temperature control system for electrical appliances based on data analysis, including a data integration module, a local model module, a cloud aggregation module, and a real-time regulation module; The data integration module is used to collect environmental data, scan the spatial thermal field distribution, calculate the initial thermal field uniformity index, and integrate it into a local data set through the edge controller; The local model module is used to train a local thermodynamic model, generate an initial symbolic equation, calculate the fitting error, obtain equation coefficients and performance metrics, perform data augmentation, and upload the results to the cloud. The cloud aggregation module is used to perform federated symbolic equation aggregation in the cloud, output a globally optimized equation, send it to the edge controller, replace the local thermodynamic model, and adjust the thermal conductivity of the finite element mesh. The real-time regulation module is used to build a digital twin environment in the edge controller, perform real-time thermal field regulation, trigger local symbolic regression retraining based on the collected feedback, and upload the retrained coefficients to the cloud to participate in the next round of federated optimization.

[0013] In a third aspect, the present invention provides a computer device, including a memory and a processor, where the memory stores a computer program, and: when the computer program is executed by the processor, any step of the method for intelligent temperature control of electrical appliances based on data analysis as described in the first aspect of the present invention is implemented.

[0014] In a fourth aspect, the present invention provides a computer-readable storage medium, on which a computer program is stored, and: when the computer program is executed by the processor, any step of the method for intelligent temperature control of electrical appliances based on data analysis as described in the first aspect of the present invention is implemented.

[0015] The beneficial effects of the present invention are as follows: By collecting environmental data and constructing a local dataset, accurate perception of the spatial thermal field is achieved, providing data support for subsequent thermodynamic modeling; By training a local thermodynamic model, generating a symbolic equation and optimizing the fitting error, the temperature control strategy has stronger adaptability, and the optimized model is uploaded to the cloud; Federated symbolic equation aggregation is performed in the cloud to generate a globally optimized equation and used to dynamically adjust the thermal conductivity of the finite element mesh, thereby improving the thermal field uniformity and temperature control accuracy; A digital twin environment is built in the edge controller, local symbolic regression retraining is triggered based on real-time data feedback, and optimized parameters are uploaded to continuously iterate the global model, thereby achieving precise, efficient, and adaptive intelligent temperature control, improving energy efficiency, and optimizing the user experience. Description of the Drawings

[0016] To more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0017] Figure 1 It is an architecture diagram of an intelligent temperature control system for electrical appliances based on data analysis in Embodiment 1.

[0018] Figure 2It is a flowchart of an intelligent temperature control method for electrical appliances based on data analysis in Embodiment 1.

[0019] Figure 3 It is a schematic diagram of the aggregation of federal symbolic equations in Embodiment 1.

[0020] Figure 4 It is a schematic diagram of digital twin regulation in Embodiment 1. Detailed implementation manners

[0021] To make the above objects, features, and advantages of the present invention more obvious and understandable, the following will describe the detailed implementation manners of the present invention in conjunction with the drawings in the specification.

[0022] In the following description, many specific details are set forth in order to fully understand the present invention. However, the present invention can also be implemented in other ways different from those described herein. Those skilled in the art can make similar extensions without departing from the connotation of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0023] Secondly, the so-called "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that can be included in at least one implementation manner of the present invention. The "in one embodiment" that appears in different places in this specification does not all refer to the same embodiment, nor is it a separate or alternative embodiment that is mutually exclusive with other embodiments.

[0024] Embodiment 1, referring to Figures 1 to 4 , which is the first embodiment of the present invention. This embodiment provides an intelligent temperature control method for electrical appliances based on data analysis, including the following steps: S1: Collect environmental data, scan the spatial thermal field distribution, calculate the initial thermal field uniformity index, and integrate it into a local data set through an edge controller.

[0025] Specifically, it includes the following steps: S1.1: Deploy a sensor array.

[0026] Specifically, deploy an infrared thermal imaging sensor array inside the target space (such as a room, equipment cabin), with the sensor spacing distributed in a grid pattern of 0.5m×0.5m (deploy one layer every 1m in the vertical direction) to ensure three-dimensional space coverage.

[0027] Each grid point is configured with a multi-modal sensor node, integrating: A temperature sensor (accuracy ±0.2°C, range -20°C to 80°C).

[0028] A humidity sensor (accuracy ±3%RH, range 0 to 100%RH).

[0029] Wind speed sensor (ultrasonic type, measuring range 0~10m / s, resolution 0.1m / s).

[0030] The multi-modal sensor node communicates with the edge controller through the Zigbee 3.0 wireless protocol, with a transmission period of 100ms, and supports dynamic adjustment of the sampling frequency (±20ms).

[0031] During the first deployment, cross-calibration is carried out in a constant temperature and humidity chamber (25℃±0.1℃, 50%RH±1%). Apply the five-point calibration method (0%, 25%, 50%, 75%, 100% range points) to each sensor node and record the output curve. Calibration is performed by the least squares method, and the sensor error after calibration is controlled within ±0.1℃ (temperature), ±2%RH (humidity), ±0.05m / s (wind speed).

[0032] S1.2: Collect environmental data.

[0033] Specifically, the IEEE 1588 PTP precise time protocol is used to achieve data synchronization. The master clock is synchronized with the clocks of each sensor node, and the time deviation ≤1ms.

[0034] Attach a quadruple identifier to each data packet, including spatial coordinates, sensor ID, timestamp, and data type.

[0035] The collected environmental data includes temperature, humidity, and wind speed.

[0036] The MQTT protocol is used for data transmission, and the QoS level is set to 1 (at least once delivery), and the bandwidth occupancy rate ≤10%.

[0037] S1.3: Calculate the initial thermal field uniformity index.

[0038] Specifically, perform Gaussian filtering on the temperature data to eliminate sensor noise, and calculate the spatial average temperature based on the temperature data.

[0039] Calculate the initial thermal field uniformity index, and the expression is: ; Where, is the initial thermal field uniformity index, is the position weight (the weight of the area around the device is set to 2.0, and the ordinary area is 1.0), is the spatial coordinate at the temperature data, is the spatial average temperature, is the maximum temperature in the space at the current moment, is the minimum temperature in the space at the current moment, is the three-dimensional spatial coordinate.

[0040] S1.4: Process abnormal environmental data.

[0041] It should be noted that the sliding window is used to calculate the temperature change rate for mutation detection. If the temperature change rate is greater than 2°C / s, it is marked as a mutation anomaly.

[0042] Based on the Mahalanobis distance to judge data outliers, if the Mahalanobis distance is greater than 3.0, it is determined as a sensor failure and the node self-check is triggered.

[0043] For continuous missing data (within 3 seconds), the spatio-temporal Kriging interpolation method is used to fill in the missing values. If it is more than three seconds, it is determined as a sensor failure and the node self-check is triggered.

[0044] For the data marked as mutation anomalies, the exponentially weighted moving average is used for smoothing processing to replace the abnormal values.

[0045] S1.5: Integrate the local dataset.

[0046] It should be understood that the processed data is formatted and saved in JSON format. The LZ77 algorithm is used for data compression, and the AES-256-GCM mode is used for encryption. The finally generated local dataset is stored in the edge node, retaining the historical data of the last 30 days.

[0047] Preferably, through the precise sensor array layout, high-precision calibration, and multi-modal data fusion, high-quality environmental data acquisition is achieved. The IEEE 1588 PTP protocol is used to ensure clock synchronization, and the MQTT protocol is used to optimize data transmission, improving data consistency and transmission efficiency. Through technologies such as Gaussian filtering for noise reduction, mutation detection, Mahalanobis distance outlier identification, and missing value filling, the accuracy and reliability of the acquisition are improved. At the same time, the thermal field uniformity index is used to quantify the spatial temperature distribution, providing data support for thermal management optimization. The data storage uses JSON formatting, LZ77 compression, and AES-256-GCM encryption to ensure efficient and secure storage, and retains the historical data of the last 30 days for trend analysis and fault tracing.

[0048] S2: Train the local thermodynamic model, generate the initial symbolic equation and calculate the fitting error, obtain the equation coefficients and performance indicators, perform data augmentation, and upload to the cloud.

[0049] Specifically, it includes the following steps: S2.1: Preprocess the local dataset.

[0050] Specifically, environmental features are extracted from the local dataset, including calculating the temperature field gradient through the central difference method, calculating the humidity change rate, and the wind speed direction angle.

[0051] The extracted environmental features are mapped to the interval [0, 1] using Min - Max normalization, and the logarithmic transformation is performed on the temperature field gradient to suppress the dimensional difference.

[0052] S2.2: Use the polynomial regression algorithm to construct the local thermodynamic model.

[0053] Specifically, a third - order polynomial regression is used to construct the local thermodynamic model, and the expression is: ; Among them, is the predicted output of the local thermodynamic model, is the intercept term, is the coefficient of the first - order term, is the coefficient of the interaction term, is the coefficient of the third - order term, is the input environmental feature variable, is another input environmental feature variable, is the number of environmental features, is the index of the number of environmental features, and the value range is , is greater than or equal to and another index of the number of environmental features, and the value range is .

[0054] The L2 regularization is added to the loss function, and the L - BFGS algorithm is used to solve the optimal coefficients. The maximum number of iterations is set to 500 times, and the convergence threshold is 1e - 6.

[0055] S2.3: Calculate the fitting error.

[0056] It should be noted that the fitting error includes the mean absolute error (MAE), the root mean square error (RMSE), and the coefficient of determination ( ).

[0057] The training set (70%), the validation set (15%), and the test set (15%) are divided. If the RMSE of the validation set > 1.2 × the RMSE of the training set, it is determined as overfitting, and the polynomial order is reduced and retrained.

[0058] S2.4: Obtain the equation coefficients and performance indicators.

[0059] Specifically, the coefficient matrix and the intercept term are extracted from the local thermodynamic model and stored in JSON format in the order of environmental features.

[0060] At the same time, for index screening, only the parameters of the local thermodynamic model with <0.85 and MAE < 0.05 are retained, and the coefficient accuracy is reserved to 4 decimal places.

[0061] S2.5: Perform data augmentation on the local dataset.

[0062] It should be noted that data augmentation refers to adding anisotropic Gaussian noise to each sample in the local dataset to generate an augmented dataset that is 10 times the amount of the original data.

[0063] S2.6: Package the equation coefficients, performance metrics, and augmented dataset and upload them to the cloud.

[0064] Specifically, package the equation coefficients, performance metrics, and augmented dataset into the Protobuf binary format, including: local thermodynamic model metadata (training time, edge node ID, hardware version), equation coefficient matrix, performance metrics, and augmented dataset.

[0065] Use the TLS 1.3 protocol for encrypted transmission. The certificate is issued by the cloud CA, and the data packet is appended with a SHA-256 hash checksum. The transmission frequency is set to once every 6 hours, and the bandwidth peak is limited to 50 Mbps.

[0066] Preferably, improve the quality of the local dataset through feature extraction, normalization, and regularization, and train the local thermodynamic model using the third-order polynomial regression combined with the L-BFGS optimization algorithm. Dynamically adjust the order to prevent overfitting and ensure high precision and generalization ability. Use anisotropic Gaussian noise for data augmentation to improve the adaptability of the local thermodynamic model. The data storage uses the Protobuf format, the transmission uses TLS 1.3 encryption, and a SHA-256 hash checksum is appended to ensure efficiency and security.

[0067] S3: Perform federated symbolic equation aggregation in the cloud, output the global optimized equation, send it to the edge controller, replace the local thermodynamic model, and adjust the thermal conductivity of the finite element mesh.

[0068] Specifically, it includes the following steps: S3.1: Select the non-dominated solution set through the Pareto front screening algorithm.

[0069] Specifically, the input is the set of local equation coefficients uploaded by each edge node, and each set of equation coefficients is associated with performance metrics.

[0070] The optimization objective is to maximize , minimize MAE, and minimize the equation complexity (number of terms).

[0071] The dominance relationship is determined as: Equation 1 dominates Equation 2 if and only if: ; And at least one objective is strictly better.

[0072] Furthermore, the NSGA-II algorithm is specifically implemented as follows: perform fast non-dominated sorting, stratify all equations according to the dominance relationship to generate non-dominated levels; calculate the crowding degree, sort the equations at the same level according to the distance in the objective space, and retain the diversity; according to the screening results of the crowding degree, select the top 50% of the equations to form the non-dominated solution set.

[0073] S3.2: Perform symbol-level crossover operation.

[0074] Specifically, select two parent equations from the non-dominated solution set, exchange the sub-expressions containing the temperature difference term and the angular trigonometric function term to generate new offspring equations.

[0075] At the same time, eliminate invalid crossover results, such as terms with no physical meaning .

[0076] S3.3: Apply Gaussian coefficient perturbation, perform equation population federated iteration, and output the globally optimized equation.

[0077] Specifically, apply adaptive Gaussian perturbation to the coefficients of the offspring equations. The expression is: ; where is the coefficient of the perturbed offspring equation, is the original coefficient of the offspring equation, is the Gaussian noise term, is the Gaussian distribution, is the adaptively adjusted noise variance.

[0078] If the is improved after perturbation, amplify the perturbation intensity: .

[0079] The termination condition of the equation population federated iteration is set to reach the maximum number of iterations (50 times) or the global equation is continuously improved by <0.1% for 5 consecutive times on the validation set.

[0080] Retain the top 10% of the optimal equations in each round to avoid performance degradation. Finally, select the equation with the highest value and complexity ≤ 5 terms on the Pareto front as the global equation. The equation with the highest value and complexity ≤ 5 terms on the Pareto front is selected as the global equation.

[0081] S3.4: Send the globally optimized equation to the edge controller to replace the local thermodynamic model.

[0082] Specifically, encode the globally optimized equation into the JSON-LD format and send it to the edge controller through the HTTP / 2 protocol. The data packet is attached with an HMAC-SHA256 signature to prevent tampering.

[0083] The edge controller loads the globally optimized equation into the memory buffer for verification ≥0.85. Atomically replace the local thermodynamic model during operation, and keep the old version for 24 hours for rollback.

[0084] S3.5: Adjust the thermal conductivity of the finite element mesh.

[0085] Specifically, calculate the sensitivity of the temperature at each grid point to the initial thermal field uniformity index according to the global optimization equation. The expression is: ; where is the thermal field uniformity sensitivity at grid point , is the partial derivative of the thermal field uniformity with respect to the grid point temperature, is the partial derivative of the thermal field uniformity with respect to the environmental characteristic variable , is the environmental characteristic variable with respect to temperature partial derivative.

[0086] The dynamic adjustment expression of the thermal conductivity is: ; where is the thermal conductivity of the updated grid point , is the current thermal conductivity, is the sensitivity coefficient ( ), which controls the adjustment amplitude of the thermal conductivity to prevent instability caused by excessive changes, is the set maximum thermal conductivity, which is used to limit the growth range of the thermal conductivity and prevent numerical values from being too large and affecting the calculation stability.

[0087] Update the thermal conductivity in real time through the API of the finite element analysis software (such as COMSOL), and iteratively solve the heat conduction equation until convergence (residual < 1e-4).

[0088] Preferably, aggregate and optimize the thermodynamic model through the federated symbolic equation to improve the temperature control performance. Select a set of equations with high precision, low error, and low complexity through the Pareto front screening, and enhance the generalization ability through symbolic-level crossover. In the Gaussian coefficient perturbation optimization, adaptively adjust the equation coefficients to improve the fitting degree and prevent model degradation. The optimized global equation is securely sent to the edge controller in JSON-LD format through HTTP / 2 + HMAC-SHA256 to replace the local model, and a rollback mechanism is provided to ensure stability. Optimize the thermal conductivity of the finite element mesh based on the global equation, calculate the thermal field uniformity sensitivity, and dynamically adjust the thermal conductivity through the COMSOL API to accelerate the thermal equilibrium and improve the energy efficiency.

[0089] S4: Build a digital twin environment in the edge controller for real-time thermal field regulation. Trigger local symbolic regression retraining based on the collected feedback, and upload the retraining coefficients to the cloud to participate in the next round of federated optimization.

[0090] Specifically, it includes the following steps: S4.1: Build a digital twin environment in the edge controller.

[0091] Specifically, generate a three-dimensional mesh model through the Open3D framework, and the mesh resolution is consistent with the physical space (0.5m × 0.5m × 1m).

[0092] Map the updated finite element mesh thermal conductivity to the digital twin model, and realize data synchronization between the physical space and the digital twin through the OPC UA protocol. The update frequency of the temperature field data is 100ms, and the update frequency of the wind speed and humidity data is 500ms. At the same time, calculate the mean square error between the digital twin predicted temperature and the actual measured value, and require MSE < 0.1°C.

[0093] S4.2: Update the policy network through the PPO algorithm.

[0094] Specifically, define the state space, including temperature, temperature field gradient, wind speed, temperature change rate, and device real-time power (collected from the device interface).

[0095] Define the action space, including temperature adjustment, wind speed adjustment, and finite element mesh thermal conductivity adjustment.

[0096] Furthermore, the architecture of the PPO algorithm policy network is as follows: input layer: 8-dimensional state → hidden layer: 256 neurons (ReLU) → output layer: 3-dimensional action mean and log standard deviation.

[0097] The discount factor is set to 0.99, the GAE parameter is set to 0.95, the policy update step size is set to 0.2, and the learning rate is set to .

[0098] Collect 2000-step interaction data in each round, with a mini-batch size of 64 and 3 epochs of iteration. When the average reward fluctuates < 1% continuously for 10 rounds, it is determined to converge.

[0099] S4.3: Generate a control signal according to the policy network for real-time thermal field regulation.

[0100] Specifically, the policy network outputs an action probability distribution, and samples to obtain a control instruction. Drive the stepper motor to adjust the air outlet angle, requiring a resolution of up to 0.5°, and the corresponding time ≤ 50ms.

[0101] Control the fan speed through the frequency converter with an accuracy of ±0.05m / s.

[0102] Adjust the local thermal conductivity, the driving voltage of the piezoelectric material, and the linear relationship between voltage and thermal conductivity.

[0103] After each action is executed, detect the change. If the change rate is greater than 0.02 / s, it is determined as an effective regulation. If the regulation is ineffective for 3 consecutive times, the maximum wind speed is forced for emergency refrigeration.

[0104] S4.4: Perform feedback acquisition according to the regulation result, trigger local symbolic regression retraining, and generate retraining coefficients.

[0105] It should be understood that retraining is started when the following conditions are met: ; Among them, is the number of sampling points within the trigger judgment window, is the sampling point index, is the predicted thermal field uniformity, is the upper limit of the prediction deviation.

[0106] Perform noise injection on historical environmental data, and the expression is: ; Among them, is the enhanced temperature data for amplifying the training set, is the original temperature data, is the noise amplitude coefficient, taking 0.5 to avoid excessive distortion, is the space axis direction length, is a Gaussian distribution random noise with a mean of 0 and a standard deviation of 0.2.

[0107] Use a deep symbolic regression network for retraining, and add equation complexity penalty to the loss function. The expression is: ; Among them, is the total loss function, is the mean square error, is the number of mathematical terms of the symbolic equation.

[0108] After training, new coefficients are obtained, and it is required that .

[0109] S4.5: Upload the retraining coefficients to the cloud to participate in the next round of federated optimization.

[0110] Preferably, real-time thermal field regulation is achieved through a digital twin environment to improve the regulation accuracy. A physical-virtual synchronization mechanism is established based on Open3D and OPCUA, and the prediction error is constrained. The PPO reinforcement learning is used to optimize the control strategy, adjust the temperature, wind speed, and thermal conductivity, and regulate the thermal field with the optimal strategy to ensure an angular accuracy of 0.5°, a response time of 50 ms, and a wind speed control of ±0.05 m / s. If the prediction deviation of the thermal field uniformity exceeds the threshold, local symbolic regression retraining is triggered, the training set is expanded through noise enhancement, and the symbolic equation is optimized. The retrained coefficients are uploaded to the cloud to participate in the next round of federated optimization, realizing the self-evolution of edge intelligence and improving the adaptability and stability of thermal field regulation.

[0111] This embodiment also provides an intelligent temperature control system for electrical appliances based on data analysis, including: a data integration module, a local model module, a cloud aggregation module, and a real-time regulation module; The data integration module is used to collect environmental data, scan the spatial thermal field distribution, calculate the initial thermal field uniformity index, and integrate it into a local data set through an edge controller; The local model module is used to train a local thermodynamics model, generate an initial symbolic equation and calculate the fitting error, obtain the equation coefficients and performance indicators, perform data augmentation, and upload them to the cloud; The cloud aggregation module is used to perform federated symbolic equation aggregation in the cloud, output a globally optimized equation, send it to the edge controller, replace the local thermodynamics model, and adjust the thermal conductivity of the finite element mesh; The real-time regulation module is used to build a digital twin environment in the edge controller, perform real-time thermal field regulation, trigger local symbolic regression retraining according to the collected feedback, and upload the retrained coefficients to the cloud to participate in the next round of federated optimization.

[0112] This embodiment also provides a computer device applicable to the case of the intelligent temperature control method for electrical appliances based on data analysis, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the intelligent temperature control method for electrical appliances based on data analysis as proposed in the above embodiment.

[0113] The computer device may be a terminal, which includes a processor, a memory, a communication interface, a display screen, and an input device connected via a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The communication interface of the computer device is used to communicate with external terminals in a wired or wireless manner, and the wireless manner can be implemented through WIFI, carrier networks, NFC (Near Field Communication), or other technologies. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device can be a touch layer covering the display screen, or buttons, trackballs, or touchpads provided on the outer shell of the computer device, or an external keyboard, touchpad, or mouse, etc.

[0114] This embodiment also provides a storage medium, on which a computer program is stored. When the program is executed by a processor, it implements the method for intelligent temperature control of electrical appliances based on data analysis proposed in the above embodiment; the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM for short), Electrically Erasable Programmable Read-Only Memory (EEPROM for short), Erasable Programmable Read-Only Memory (EPROM for short), Programmable Read-Only Memory (PROM for short), Read-Only Memory (ROM for short), magnetic memory, flash memory, magnetic disks, or optical discs.

[0115] In summary, the present invention realizes precise perception of the spatial thermal field by collecting environmental data and constructing a local data set, providing data support for subsequent thermodynamic modeling; by training a local thermodynamic model, generating symbolic equations and optimizing the fitting error, making the temperature control strategy more adaptable, and uploading the optimized model to the cloud; aggregating the federated symbolic equations in the cloud to generate a globally optimized equation and using it to dynamically adjust the thermal conductivity of the finite element mesh, thereby improving the thermal field uniformity and temperature control accuracy; building a digital twin environment in the edge controller, triggering local symbolic regression retraining based on real-time data feedback, and uploading optimized parameters to continuously iterate the global model, so as to achieve precise, efficient, and adaptive intelligent temperature control, improve energy efficiency, and optimize the user experience.

[0116] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered by the scope of the claims of the present invention.

Claims

1. An intelligent temperature control method for electrical appliances based on data analysis, characterized in that: include, Collect environmental data, scan the spatial thermal field distribution, calculate the initial thermal field uniformity index, and integrate it into a local data set through the edge controller; Train the local thermodynamic model, generate the initial symbolic equation and calculate the fitting error, obtain the equation coefficients and performance indicators, perform data enhancement, and upload to the cloud; Federated symbolic equation aggregation is performed in the cloud, and global optimization equations are output and sent to the edge controller to replace the local thermodynamic model and adjust the thermal conductivity of the finite element mesh. Build a digital twin environment on the edge controller to perform real-time thermal field control, trigger local symbolic regression retraining based on collected feedback, and upload the retraining coefficients to the cloud to participate in the next round of federated optimization.

2. The method for intelligent temperature control of an electrical appliance based on data analysis as claimed in claim 1, characterized in that: The federated symbolic equation aggregation refers to selecting a non-dominated solution set through a Pareto front screening algorithm, performing a symbolic crossover operation and applying a Gaussian coefficient perturbation, and performing a federated iterative evolution of an equation population.

3. The method for intelligent temperature control of an electrical appliance based on data analysis as claimed in claim 2, characterized in that: Adjusting the thermal conductivity of the finite element grid refers to dynamically adjusting the thermal conductivity of each grid according to the trend of the thermal field uniformity change output by the global optimization equation, wherein the adjustment range of the thermal conductivity is proportional to the sensitivity of the thermal field uniformity to the local temperature, and the maximum adjustment range is controlled by a preset sensitivity coefficient.

4. The method for intelligent temperature control of an electrical appliance based on data analysis as claimed in claim 1, characterized in that: The training of the local thermodynamic model refers to performing data cleaning and feature normalization on the local data set, building the local thermodynamic model using a polynomial regression algorithm, and performing training, verification and optimization through the local data set.

5. The method for intelligent temperature control of an electrical appliance based on data analysis as claimed in claim 4, characterized in that: The local symbolic regression retraining means that when the actual value of the thermal field uniformity calculated in real time deviates from the predicted value of the local thermodynamic model by more than a retraining threshold, Gaussian noise is added to the historical environmental data to generate an enhanced training set, and the symbolic regression algorithm is re-run to obtain the retrained equation coefficients.

6. The method for intelligent temperature control of an electrical appliance based on data analysis as claimed in claim 3, characterized in that: The digital twin environment is constructed to perform real-time thermal field control. The specific steps are: Define the state space, action space and reward function, and use the proximal policy optimization algorithm to train the policy network; After the training converges, the spatial thermal field distribution is scanned in real time, and the optimal action is selected according to the current state; The stepper motor is driven to adjust the outlet angle and the frequency converter to adjust the wind speed, and a pulse signal is sent to the piezoelectric material controller to change the local thermal conductivity.

7. The method for intelligent temperature control of an electrical appliance based on data analysis as claimed in claim 2, characterized in that: The symbolic crossover operation refers to selecting two parent equations from the non-dominated solution set, exchanging their sub-expressions containing temperature difference terms and angle trigonometric function terms, generating new equations, and applying random perturbations to the coefficients of the new equations.

8. An intelligent temperature control system for an electrical appliance based on data analysis, based on the intelligent temperature control method for an electrical appliance based on data analysis according to any one of claims 1 to 7, characterized in that: Including data integration module, local model module, cloud aggregation module and real-time control module; The data integration module is used to collect environmental data, scan the spatial thermal field distribution, calculate the initial thermal field uniformity index, and integrate it into a local data set through the edge controller; The local model module is used to train the local thermodynamic model, generate the initial symbolic equation and calculate the fitting error, obtain the equation coefficients and performance indicators, perform data enhancement, and upload to the cloud; The cloud aggregation module is used to aggregate the federated symbolic equations in the cloud, output the global optimization equations, send them to the edge controller, replace the local thermodynamic model, and adjust the thermal conductivity of the finite element mesh; The real-time control module is used to build a digital twin environment on the edge controller, perform real-time thermal field control, trigger local symbolic regression retraining based on collected feedback, and upload the retraining coefficients to the cloud to participate in the next round of federated optimization.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the electrical appliance intelligent temperature control method based on data analysis as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the electrical appliance intelligent temperature control method based on data analysis as described in any one of claims 1 to 7 are implemented.

Images