A precise heating control method, device, equipment and storage medium for an electric stove

Abstract

The present invention discloses a precise heating control method, device, equipment and storage medium for an electric stove, which relates to the field of electric stove heating control. The method includes: constructing a target input vector according to the obtained cookware type, cookware material, the temperature of the center point of the cookware at the current moment, and the working state information of the plasma electrode of the electric stove at the current moment; inputting the target input vector into a temperature prediction model trained by a historical sample data set, and predicting the target prediction temperature of the target input vector at the next moment through the temperature prediction model; adjusting the power output strategy of the plasma electrode of the electric stove according to the relationship between the target prediction temperature and the preset working temperature. The present invention can precisely control the power output distribution of the plasma electrode according to the cookware material, cookware type and heating requirements.

Description

Technical Field

[0001] The present invention relates to the technical field of electric stove heating control, and particularly to a precise heating control method, device, equipment and storage medium for an electric stove. Background Art

[0002] An electric stove is a kitchen appliance that converts electrical energy into heat energy through plasma technology to achieve open-flame cooking. Compared with traditional gas stoves, an electric stove does not require gas and burners, and only needs electricity to generate heat energy.

[0003] The heating method of traditional electric stoves is usually fixed power distribution. By setting multiple different firepower levels to control the size of the heat, that is, different output powers of the plasma electrodes are adapted to different firepower levels. For example: select the low fire level when stewing and cooking on low heat, select the medium fire level when stir-frying and steaming general ingredients, and select the high fire level when quickly heating and boiling water. However, the traditional heating control method cannot accurately adjust according to the pot material, pot type and heating requirements, resulting in heat waste or uneven heating.

[0004] Therefore, we need a precise heating control solution for electric stoves to accurately control the power distribution of the plasma electrodes of the electric stove by combining the material of the pot, the type of the pot and the real-time temperature of the pot. Summary of the Invention

[0005] The present invention provides a precise heating control method, device, equipment and storage medium for an electric stove, which solves the problem that the traditional heating control method cannot accurately adjust the power distribution of the plasma electrodes according to the pot material, pot type and heating requirements.

[0006] To achieve the above object, the present invention adopts the following technical solutions:

[0007] In a first aspect, the present invention provides a precise heating control method for an electric stove, and the method includes:

[0008] Construct a target input vector according to the obtained pot type, pot material, temperature of the center point of the pot at the current moment and the working state information of the plasma electrodes of the electric stove at the current moment set on the electric stove; the working state information includes the power output levels of all the plasma electrodes of the electric stove;

[0009] Input the target input vector into a temperature prediction model trained by a historical sample data set, and predict the target predicted temperature at the next moment of the target input vector through the temperature prediction model; the historical sample data set includes historical input vectors and the actual temperatures at the next moment corresponding to the historical input vectors;

[0010] Adjust the power output strategy of the plasma electrode of the electric cooker according to the relationship between the predicted target temperature and the preset operating temperature.

[0011] In a possible implementation, before predicting the target predicted temperature of the target input vector at the next moment through the temperature prediction model, the method further includes:

[0012] Conduct temperature measurement experiments on all combinations of the working state information of all plasma electrodes of the electric cooker, the cookware material information, and the cookware type information, obtain the temperature at the center point of the cookware at the current moment for each combination, and the actual temperature at the center point of the cookware at the next moment;

[0013] Construct a historical input vector for each combination according to the cookware type, cookware material, the temperature at the center point of the cookware at the current moment, and the working state information of the plasma electrode at the current moment of each combination;

[0014] Construct a historical sample data set according to multiple historical input vectors and the actual temperature at the center point of the cookware at the next moment corresponding to each historical input vector;

[0015] Train the original temperature prediction model through the historical sample data set until the loss value between the predicted temperature and the actual temperature output by the temperature prediction model no longer decreases, and obtain the trained temperature prediction model.

[0016] In a possible implementation, the loss value is calculated through the comprehensive loss function of the temperature prediction model;

[0017] The comprehensive loss function is the sum of the temperature error loss and the energy consumption loss;

[0018] For a sample group in the historical sample data set, the temperature error loss is the mean square error between the predicted temperature and the actual temperature, and the energy consumption loss is determined according to the power output gear of all plasma electrodes of the electric cooker.

[0019] In a possible implementation, the loss function of the energy consumption loss is specifically:

[0020] ;

[0021] Wherein, represents the energy consumption loss, represents the weight factor of the energy consumption loss in the comprehensive loss function, represents the serial number of the current sample in the sample group, represents the total number of samples in the sample group, represents the serial number of the plasma electrode in the current sample, represents the total number of plasma electrodes, Indicates the th power output gear of the th plasma electrode in the sample.

[0022] In a possible implementation, before predicting the target predicted temperature of the target input vector at the next moment through the temperature prediction model, the method further includes: constructing an original temperature prediction model based on a multi-layer perceptron;

[0023] The original temperature prediction model includes an input layer, a first hidden layer, a second hidden layer, and an output layer connected in sequence;

[0024] The input layer is used to receive an input vector; the input vector is a historical input vector or a target input vector;

[0025] The first hidden layer and the second hidden layer are sequentially arranged between the input layer and the output layer, and the second hidden layer is connected with the input layer by residual connection; the first hidden layer and the second hidden layer are used to perform non-linear transformation and feature extraction on the received input vector;

[0026] The output layer is used to receive the information of the second hidden layer and output the predicted temperature.

[0027] In a possible implementation, when the target predicted temperature is less than the preset working temperature, according to the relationship between the target predicted temperature and the preset working temperature, adjust the power output strategy of the plasma electrode of the induction cooker, specifically including:

[0028] Combined with the working state information of the plasma electrode at the current moment, randomly turn on an unopened plasma electrode;

[0029] And / or, randomly select half of the opened plasma electrodes, and increase the power output gear of the plasma electrodes whose power output gears have not reached the highest gear by one gear.

[0030] In a possible implementation, when the target predicted temperature is greater than the preset working temperature, according to the relationship between the target predicted temperature and the preset working temperature, adjust the power output strategy of the plasma electrode of the induction cooker, specifically including:

[0031] Combined with the working state information of the plasma electrode at the current moment, randomly turn off an opened plasma electrode;

[0032] And / or, randomly select half of the opened plasma electrodes, and reduce the power output gear of the plasma electrodes by one gear.

[0033] In a second aspect, the present invention provides an induction cooker precise heating control device, and the device includes:

[0034] A data acquisition module, configured to construct a target input vector according to the acquired cookware type, cookware material, the temperature of the center point of the cookware at the current moment, and the working state information of the plasma electrodes of the induction cooker at the current moment; the working state information includes the power output gears of all the plasma electrodes of the induction cooker.

[0035] A data processing module, configured to input the target input vector into a temperature prediction model trained by a historical sample data set, and predict the target prediction temperature at the next moment of the target input vector through the temperature prediction model; the historical sample data set includes historical input vectors and the actual temperatures at the next moment corresponding to the historical input vectors.

[0036] A strategy adjustment module, configured to adjust the power output strategy of the plasma electrodes of the induction cooker according to the relationship between the target prediction temperature and the preset working temperature.

[0037] In a possible implementation manner, the precise heating control device of the induction cooker further includes a model training module. Before predicting the target prediction temperature at the next moment of the target input vector through the temperature prediction model, the model training module is configured to execute:

[0038] Conduct a temperature measurement experiment on all combinations of the working state information of all the plasma electrodes of the induction cooker, the cookware material information, and the cookware type information, and obtain the temperature of the center point of the cookware at the current moment for each combination, and the actual temperature of the center point of the cookware at the next moment.

[0039] Construct a historical input vector for each combination according to the cookware type, cookware material, the temperature of the center point of the cookware at the current moment, and the working state information of the plasma electrodes at the current moment of each combination.

[0040] Construct a historical sample data set according to multiple historical input vectors and the actual temperature of the center point of the cookware at the next moment corresponding to each historical input vector.

[0041] Train the original temperature prediction model through the historical sample data set until the loss value between the prediction temperature output by the temperature prediction model and the actual temperature no longer decreases, and obtain the trained temperature prediction model.

[0042] In a possible implementation manner, in the model training module, the loss value is configured to be calculated by a comprehensive loss function of the temperature prediction model.

[0043] The comprehensive loss function is the sum of the temperature error loss and the energy consumption loss.

[0044] For a sample group in the historical sample dataset, the temperature error loss is the mean square error between the predicted temperature and the actual temperature, and the energy consumption loss is determined according to the power output levels of all plasma electrodes of the induction cooker.

[0045] In a possible implementation, in the model training module, the loss function of the energy consumption loss is specifically configured as:

[0046] ;

[0047] Wherein, represents the energy consumption loss, represents the weight factor of the energy consumption loss in the comprehensive loss function, represents the number of the current sample in the sample group, represents the total number of samples in the sample group, represents the number of the plasma electrode in the current sample, represents the total number of plasma electrodes, represents the th power output level of the

[0048] In a possible implementation, the precise heating control device of the induction cooker further includes a model construction module. Before predicting the target predicted temperature of the target input vector at the next moment through the temperature prediction model, the model construction module is configured to execute:

[0049] Construct an original temperature prediction model based on a multi-layer perceptron;

[0050] The original temperature prediction model includes an input layer, a first hidden layer, a second hidden layer, and an output layer connected in sequence;

[0051] The input layer is used to receive the input vector; the input vector is a historical input vector or a target input vector;

[0052] The first hidden layer and the second hidden layer are sequentially arranged between the input layer and the output layer, and the second hidden layer is connected with the input layer by residual connection; the first hidden layer and the second hidden layer are used to perform non-linear transformation and feature extraction on the received input vector;

[0053] The output layer is used to receive the information of the second hidden layer and output the predicted temperature.

[0054] In a possible implementation, when the target predicted temperature is less than the preset working temperature, the policy adjustment module is configured to execute:

[0055] Combined with the working state information of the plasma electrode at the current moment, randomly turn on an unopened plasma electrode;

[0056] And / or, randomly select half of the turned-on plasma electrodes and increase the power output level of the plasma electrodes whose power output level has not reached the highest level by one level.

[0057] In a possible implementation manner, when the target predicted temperature is greater than the preset working temperature, the policy adjustment module is configured to execute:

[0058] Combined with the working state information of the plasma electrode at the current moment, randomly turn off a turned-on plasma electrode;

[0059] And / or, randomly select half of the turned-on plasma electrodes and decrease the power output level of the plasma electrodes by one level.

[0060] In a third aspect, the present invention provides a device using plasma heating, including:

[0061] A controller;

[0062] A memory for storing executable instructions of the controller;

[0063] Wherein, the controller is configured to execute the instructions to implement the precise heating control method of the induction cooker as described in any one of the above.

[0064] In a fourth aspect, the present invention provides a computer-readable storage medium, in which at least one instruction, at least one segment of program, code set or instruction set is stored, and the at least one instruction, the at least one segment of program, the code set or instruction set is loaded and executed by a processor to implement the precise heating control method of the induction cooker as described in any one of the above.

[0065] When the precise heating control method for an electric stove provided by an embodiment of the present invention is actually applied, first, the type of cookware, the material of the cookware, the temperature of the center point of the cookware at the current moment, and the power output gear of each group of plasma electrodes of the electric stove at the current moment are collected, and a target input vector is constructed according to the information of the collected electric stove. Among them, as the gear increases, the output power of the plasma electrode increases. Then, the target input vector is input into the trained temperature prediction model, which is used to predict the target prediction temperature of the cookware at the next moment according to the target input vector. The temperature prediction model is trained by the historical input vector and the actual temperature at the next moment corresponding to the historical input vector. Finally, according to the relationship between the target prediction temperature output by the temperature prediction model and the preset working temperature required by the user, the power output strategy of the plasma electrodes of the electric stove is adjusted. The present invention can accurately control the power output distribution of the plasma electrodes according to the cookware material, cookware type and the heating requirements of the user, so as to dynamically adjust the temperature of the cookware, keep the temperature of the cookware dynamically at the preset working temperature, dynamically control the power output and power distribution of multiple groups of plasma heads of the electric stove, make the heat distribution more uniform, improve the heating efficiency and reduce the waste of thermal energy. Description of the Drawings

[0066] Figure 1 It is a flowchart of the steps of a precise heating control method for an electric stove provided by an embodiment of the present invention;

[0067] Figure 2 It is a temperature curve of the temperature of the center point of the cookware corresponding to a precise heating control method for an electric stove provided by an embodiment of the present invention;

[0068] Figure 3 It is a structural block diagram of a precise heating control device for an electric stove provided by an embodiment of the present invention. Detailed Embodiments

[0069] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the 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 of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0070] Hereinafter, the terms "first" and "second" are for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, features defined with "first" and "second" may explicitly or implicitly include one or more such features. In the description of the embodiments of the present disclosure, unless otherwise stated, "a plurality of" means two or more. Additionally, the use of "based on" or "according to" is meant to be open and inclusive, because a process, step, calculation, or other action "based on" or "according to" one or more of the stated conditions or values may in practice be based on additional conditions or values beyond those stated.

[0071] To solve the problem that the traditional heating control method cannot accurately regulate the power distribution of the plasma electrodes according to the cookware material, cookware type, and user heating requirements, the embodiments of the present invention provide a precise heating control method, device, equipment, and storage medium for an electric stove.

[0072] As Figure 1 shown, in a first aspect, the embodiments of the present invention provide a precise heating control method for an electric stove, and the method includes:

[0073] Step 101, construct a target input vector according to the obtained cookware type, cookware material, temperature of the center point of the cookware at the current moment, and the working state information of the plasma electrodes of the electric stove at the current moment.

[0074] Among them, the working state information includes the power output gears of all the plasma electrodes of the electric stove.

[0075] Specifically, each electric stove includes multiple groups of plasma electrodes. Number all the plasma electrodes, divide the maximum output power ladder of each group of plasma electrodes into multiple power levels, and correspond each power level to a power output gear.

[0076] In this embodiment, the electric stove includes 18 groups of plasma electrodes, and the 18 groups of plasma electrodes are numbered sequentially. Then, the maximum output power ladder of each group of plasma electrodes is divided into 4 levels. When the maximum output power of each group of plasma electrodes is U, the selectable output powers of each group of plasma electrodes after grading include 0, U / 3, 2U / 3, and U, and these four output powers correspond to the 0th gear, 1st gear, 2nd gear, and 3rd gear of the plasma electrode power output gear respectively.

[0077] The working state information is an 18-dimensional vector, specifically denoted as:

[0078] ;

[0079] Among them, is used to represent the plasma electrode power output gear, and its value can be 0, 1, 2, or 3; The subscript is used to represent the number of the plasma electrode.

[0080] For different types and materials of cookware, the temperature changes during heating are not the same.

[0081] Among them, the types of cookware include soup pots, frying pans, sauté pans, and steamers, etc., and the materials of cookware include iron, aluminum, and stainless steel, etc.

[0082] In this embodiment, the material of the cookware is represented by a three-dimensional vector:

[0083] ;

[0084] Among them, , , respectively represent different cookware materials, and their values can be 1 or 0. For example: the vector corresponding to iron is (1, 0, 0), the vector corresponding to aluminum is (0, 1, 0), and the vector corresponding to stainless steel is (0, 0, 1).

[0085] The type of cookware is represented by a four-dimensional vector:

[0086] ;

[0087] Among them, , , , respectively represent different types of cookware, and their values can be 1 or 0. For example: the vector corresponding to a soup pot is (1, 0, 0, 0), the vector corresponding to a frying pan is (0, 1, 0, 0), the vector corresponding to a sauté pan is (0, 0, 1, 0), and the vector corresponding to a steamer is (0, 0, 0, 1).

[0088] Based on this, according to the type of cookware, the material of the cookware, the vector corresponding to the power output gear of each plasma electrode of the induction cooker, and the temperature of the center point of the cookware measured by the temperature detector at the current moment, the target input vector is determined.

[0089] Step 102: Input the target input vector into the temperature prediction model trained by the historical sample data set, and predict the target prediction temperature at the next moment of the target input vector through the temperature prediction model.

[0090] Among them, the historical sample data set includes historical input vectors and the actual temperatures at the next moment corresponding to the historical input vectors.

[0091] In the embodiment of the present invention, the target input vector is composed of , , and It consists of these four groups of variables, where is an 18-dimensional vector, is a 3-dimensional vector, is a 4-dimensional vector, is a 1-dimensional vector.

[0092] Input the target input vector into the trained temperature prediction model for prediction, and the target input vector at the next moment can be accurately output as the target predicted temperature.

[0093] Step 103: Adjust the power output strategy of the plasma electrode of the induction cooker according to the relationship between the target predicted temperature and the preset working temperature.

[0094] Specifically, the preset working temperature can be set by the function buttons of the induction cooker. In actual applications, under the heating of the plasma electrode, the temperature of the cookware continuously rises and finally reaches the preset working temperature, and the temperature needs to be controlled within the preset working temperature.

[0095] That is to say, the temperature control of the induction cooker is a dynamic balance process. When the target predicted temperature is less than the preset working temperature, it is necessary to increase the output power of some plasma electrodes or turn on several more groups of plasma electrodes to make the temperature of the induction cooker gradually approach the preset working temperature during actual operation; when the target predicted temperature is greater than the preset working temperature, it is necessary to reduce the output power of the plasma electrodes or turn off some plasma electrodes to make the temperature of the induction cooker gradually approach the preset working temperature during actual operation, and finally achieve dynamic balance, so as to realize precise control of the heating of the cookware.

[0096] As Figure 2 shown, the temperature at the center point of the cookware gradually rises from the initial room temperature state. When the temperature at the center point of the cookware reaches near the preset working temperature T0, through the precise heating control method of the induction cooker of the present invention, the output power of the plasma electrode of the induction cooker is dynamically adjusted to dynamically maintain the temperature at the center point of the cookware near the preset working temperature T0, so as to realize stable control of the temperature of the cookware.

[0097] When the precise heating control method for the induction cooker provided by the embodiment of the present invention is actually applied, first, the type of cookware, the material of the cookware, the temperature of the center point of the cookware at the current moment, and the power output gear of each group of plasma electrodes of the induction cooker at the current moment are collected, and a target input vector is constructed according to the information of the induction cooker collected. Among them, as the gear increases, the output power of the plasma electrode increases. Then, the target input vector is input into the trained temperature prediction model, which is used to predict the target predicted temperature of the cookware at the next moment according to the target input vector. The temperature prediction model is trained by the historical input vector and the actual temperature at the next moment corresponding to the historical input vector. Finally, according to the relationship between the target predicted temperature output by the temperature prediction model and the preset working temperature required by the user, the power output strategy of the plasma electrodes of the induction cooker is adjusted.

[0098] The present invention can accurately control the power output distribution of the plasma electrodes according to the cookware material, cookware type and the user's heating demand, so as to dynamically adjust the temperature of the cookware, keep the temperature of the cookware dynamically at the preset working temperature, dynamically control the power output and power distribution of multiple groups of plasma heads of the induction cooker, make the heat distribution more uniform, improve the heating efficiency and reduce the waste of thermal energy.

[0099] Further, before predicting the target predicted temperature of the target input vector at the next moment through the temperature prediction model, the method further includes:

[0100] Perform temperature measurement experiments on all combinations of the working state information of all plasma electrodes of the induction cooker, the cookware material information and the cookware type information, and obtain the temperature of the center point of the cookware at the current moment for each combination, and the actual temperature of the center point of the cookware at the next moment;

[0101] Construct a historical input vector for each combination according to the cookware type, cookware material, the temperature of the center point of the cookware at the current moment and the working state information of the plasma electrodes at the current moment;

[0102] Construct a historical sample data set according to multiple historical input vectors and the actual temperature of the center point of the cookware at the next moment corresponding to each historical input vector;

[0103] Train the original temperature prediction model through the historical sample data set until the loss value between the predicted temperature output by the temperature prediction model and the actual temperature no longer decreases, and obtain the trained temperature prediction model.

[0104] Among them, after vectorizing and encoding the cookware type, cookware material and the output power of each plasma electrode, temperature measurement experiments are carried out on all combinations of the above three groups of vectors, that is, the working state information of the plasma electrodes of the induction cooker at the current moment is , the cookware material is , when the type of cookware is , measure the temperature data of the center point of the cookware at the current moment with a temperature detector , and the actual temperature of the center point of the cookware at the next moment. Thus, a historical sample data set is obtained, and the historical sample data set includes multiple historical input vectors and their corresponding actual temperatures at the next moment.

[0105] When constructing the historical sample data set, the time interval between the next moment and the current moment is 10S. That is, the temperature of the center point of the cookware after 10S is related to the type of cookware, the material of the cookware, the output power levels of each plasma electrode 10S ago, and the temperature of the center point of the cookware 10S ago.

[0106] When performing model training, divide the historical sample data set into a training set, a validation set, and a test set, and the division ratios are 70%, 20%, and 10% respectively. Train the parameters and weights of the temperature prediction model through the training set; adjust the hyperparameters in the model through the validation set, such as: learning rate, regularization parameter, etc., to improve the generalization performance of the model. That is, judge whether the temperature prediction model is overfitting or underfitting through the performance of the validation set. Finally, evaluate the performance of the temperature prediction model through the test set, and check the generalization ability of the temperature prediction model for data that has never been seen. Minimize the loss function through the Adam optimizer. Calculate the gradient of the loss function with respect to each parameter through the backpropagation algorithm, and update the parameters to reduce the loss.

[0107] In the embodiment of the present invention, monitor the performance of the temperature prediction model through the validation set until the loss value between the predicted temperature and the actual temperature in the validation set no longer decreases in a continuous number of batch sizes, then stop training to avoid model overfitting, and obtain a trained temperature prediction model.

[0108] Furthermore, the loss value is calculated through the comprehensive loss function of the temperature prediction model.

[0109] The comprehensive loss function is the sum of the temperature error loss and the energy consumption loss.

[0110] For a sample group in the historical sample data set, the temperature error loss is the mean square error between the predicted temperature and the actual temperature, and the energy consumption loss is determined according to the power output levels of all plasma electrodes of the induction cooker.

[0111] Specifically, the comprehensive loss function is:

[0112] ;

[0113] Among them, represents the comprehensive loss; represents the temperature error loss, which is used to represent the error of the temperature prediction model in predicting the temperature of the cookware; Indicates the energy consumption loss.

[0114] The temperature error loss function is specifically:

[0115] ;

[0116] Among them, represents the total number of samples in the sample group, which is used to calculate the average loss and ensure that the loss value is independent of the number of samples.

[0117] represents the th temperature vector of the center point of the cookware measured at the current moment for the

[0118] th training sample. represents the predicted temperature of the

[0119] th training sample at the next moment.

[0120] ;

[0121] Among them, represents the energy consumption loss.

[0122] represents the weight factor of the energy consumption loss in the comprehensive loss function, which is used to balance the relative importance of the temperature error loss and the energy consumption loss. By adjusting the value of , the trade-off between optimizing the temperature accuracy and the energy efficiency of the temperature prediction model can be controlled.

[0123] represents the number of the current sample in the sample group, represents the total number of samples in the sample group, represents the number of the plasma electrode in the current sample, represents the total number of plasma electrodes, represents the th power output level of the th plasma electrode in the

[0124] The comprehensive loss function of the present invention consists of a temperature error loss part and an energy consumption loss part, enabling the temperature prediction model to better balance the temperature accuracy and the energy consumption during the training process.

[0125] Specifically, based on the traditional mean square error, a penalty term for energy consumption is added to form a comprehensive loss function. By introducing the loss term of energy consumption, the temperature prediction model can not only focus on the accuracy of temperature prediction during the training process, but also consider the energy efficiency of power settings, which helps to achieve the dual goals of temperature control and energy optimization, and improve the overall performance and user satisfaction of the electric stove.

[0126] Further, before predicting the target predicted temperature at the next moment of the target input vector through the temperature prediction model, the method further includes: constructing an original temperature prediction model based on a multi-layer perceptron.

[0127] The original temperature prediction model includes an input layer, a first hidden layer, a second hidden layer, and an output layer connected in sequence.

[0128] The input layer is used to receive the input vector.

[0129] Among them, the input vector is a historical input vector or a target input vector.

[0130] The first hidden layer and the second hidden layer are sequentially arranged between the input layer and the output layer, and the second hidden layer is connected with the input layer by a residual connection.

[0131] Among them, the first hidden layer and the second hidden layer are used to perform non-linear transformation and feature extraction on the received input vector, and the activation functions of the first hidden layer and the second hidden layer are both ReLU functions. Through the ReLU function, a non-linear relationship can be introduced into the hidden layer, enabling the temperature prediction model to learn complex patterns and features.

[0132] The output layer is used to receive the information of the second hidden layer and output the predicted temperature.

[0133] In this embodiment, a multi-layer perceptron is selected to construct the original temperature prediction model, and a residual connection is added between the hidden layers, so that the output of the second hidden layer not only depends on the output of the first hidden layer, but also includes a part from the input layer. The present invention allows the gradient to flow directly through the multi-layer perceptron through the residual connection, alleviating the gradient disappearance problem in the deep network of the multi-layer perceptron. This structure creates a direct "jump" path between layers, reduces information loss, and improves the training efficiency of the network. In addition, it promotes the efficient propagation of information in the network, enables the model to learn more complex feature representations, and improves the prediction performance.

[0134] Further, when the target predicted temperature is less than the preset working temperature, according to the relationship between the target predicted temperature and the preset working temperature, the power output strategy of the plasma electrode of the electric stove is adjusted, specifically including:

[0135] Combined with the working state information of the plasma electrode at the current moment, randomly turn on an unopened plasma electrode;

[0136] And / or, randomly select half of the turned-on plasma electrodes, and increase the power output level of the plasma electrodes whose power output levels have not reached the highest level by one level.

[0137] Furthermore, when the target predicted temperature is greater than the preset working temperature, adjust the power output strategy of the plasma electrodes of the induction cooker according to the relationship between the target predicted temperature and the preset working temperature, specifically including:

[0138] Combined with the working state information of the plasma electrodes at the current moment, randomly turn off one of the turned-on plasma electrodes;

[0139] And / or, randomly select half of the turned-on plasma electrodes, and lower the power output level of the plasma electrodes by one level.

[0140] In this embodiment, when the induction cooker starts to work, randomly turn on half of the plasma electrodes and randomly set the power output levels of the turned-on plasma electrodes. Input the target input vector at the current moment into the trained temperature prediction model, and the target predicted temperature at the next moment output by the temperature prediction model can be obtained. By comparing the relationship between the target predicted temperature and the preset working temperature, the power output of the plasma electrodes can be dynamically adjusted.

[0141] When the target predicted temperature is less than the preset working temperature, first randomly turn on one plasma electrode and set the level of this plasma electrode to level 2. If all plasma electrodes are in the turned-on state at this time, this operation does not need to be performed. After that, randomly select half of the turned-on plasma electrodes and increase the power output level of them by one level. If the selected plasma electrode is already at the highest level, this operation is not required.

[0142] When the target predicted temperature is greater than the preset working temperature, first randomly turn off one of the turned-on plasma electrodes. If all plasma electrodes are in the turned-off state at this time, this operation does not need to be performed. After that, randomly select half of the plasma electrodes in the turned-on state and lower the power output level of them by one level. If the selected plasma electrode is already at the lowest level, this operation is not required.

[0143] By adjusting the power output strategy through this solution, the cookware can be better stabilized within the preset working stability range after heating up from the room temperature state to the preset working stability.

[0144] As Figure 3 shown, in the second aspect, the embodiment of the present invention provides an induction cooker precise heating control device, and the device includes:

[0145] The data acquisition module 201 is configured to construct a target input vector according to the acquired cookware type, cookware material, the temperature of the center point of the cookware at the current moment, and the working state information of the plasma electrodes of the induction cooker at the current moment; the working state information includes the power output gears of all the plasma electrodes of the induction cooker.

[0146] The data processing module 202 is configured to input the target input vector into a temperature prediction model trained by a historical sample data set, and predict the target prediction temperature at the next moment of the target input vector through the temperature prediction model; the historical sample data set includes historical input vectors and the actual temperatures at the next moment corresponding to the historical input vectors.

[0147] The strategy adjustment module 203 adjusts the power output strategy of the plasma electrodes of the induction cooker according to the relationship between the target prediction temperature and the preset working temperature.

[0148] Further, the precise heating control device of the induction cooker further includes a model training module. Before predicting the target prediction temperature at the next moment of the target input vector through the temperature prediction model, the model training module is configured to perform:

[0149] Perform a temperature measurement experiment on all combinations of the working state information of all the plasma electrodes of the induction cooker, the cookware material information, and the cookware type information, and obtain the temperature of the center point of the cookware at the current moment for each combination, and the actual temperature of the center point of the cookware at the next moment.

[0150] Construct a historical input vector for each combination according to the cookware type, cookware material, the temperature of the center point of the cookware at the current moment, and the working state information of the plasma electrodes at the current moment for each combination.

[0151] Construct a historical sample data set according to multiple historical input vectors and the actual temperature of the center point of the cookware at the next moment corresponding to each historical input vector.

[0152] Train the original temperature prediction model through the historical sample data set until the loss value between the predicted temperature and the actual temperature output by the temperature prediction model no longer decreases, and obtain the trained temperature prediction model.

[0153] Further, in the model training module, the loss value is configured to be calculated by the comprehensive loss function of the temperature prediction model.

[0154] The comprehensive loss function is the sum of the temperature error loss and the energy consumption loss.

[0155] For a sample group in the historical sample data set, the temperature error loss is the mean square error between the predicted temperature and the actual temperature, and the energy consumption loss is determined according to the power output gears of all the plasma electrodes of the induction cooker.

[0156] Further, in the model training module, the loss function of the energy consumption loss is specifically configured as:

[0157] ;

[0158] where represents the energy consumption loss, represents the weight factor of the energy consumption loss in the comprehensive loss function, represents the number of the current sample in the sample group, represents the total number of samples in the sample group, represents the number of the plasma electrode in the current sample, represents the total number of plasma electrodes, represents the th power output gear of the

[0159] th plasma electrode in the

[0160] Based on the multi-layer perceptron, construct the original temperature prediction model;

[0161] The original temperature prediction model includes an input layer, a first hidden layer, a second hidden layer, and an output layer connected in sequence;

[0162] The input layer is used to receive the input vector; the input vector is a historical input vector or a target input vector;

[0163] The first hidden layer and the second hidden layer are sequentially arranged between the input layer and the output layer, and the second hidden layer is connected with the input layer by residual connection; the first hidden layer and the second hidden layer are used to perform non-linear transformation and feature extraction on the received input vector;

[0164] The output layer is used to receive the information of the second hidden layer and output the predicted temperature.

[0165] Further, when the target predicted temperature is less than the preset working temperature, the policy adjustment module 203 is configured to execute:

[0166] Combined with the working state information of the plasma electrode at the current moment, randomly turn on an unopened plasma electrode;

[0167] and / or, randomly select half of the opened plasma electrodes, and increase the power output gear of the plasma electrodes whose power output gear has not reached the highest gear by one gear.

[0168] Further, when the target predicted temperature is greater than the preset operating temperature, the policy adjustment module 203 is configured to perform:

[0169] Combined with the working state information of the plasma electrode at the current moment, randomly turn off one of the turned-on plasma electrodes;

[0170] And / or, randomly select half of the turned-on plasma electrodes and reduce the power output gear of the plasma electrodes by one gear.

[0171] The precise heating control device for an induction cooker provided by the embodiments of the present invention is used to execute the above-mentioned precise heating control method for an induction cooker, so the same effects as the above-mentioned precise heating control method for an induction cooker can be achieved.

[0172] Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working processes of the above-described systems, devices, and units can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0173] In a third aspect, the embodiments of the present invention further provide a device using plasma heating, including:

[0174] A controller;

[0175] A memory for storing executable instructions of the controller;

[0176] Wherein, the controller is configured to execute the instructions to implement the precise heating control method for an induction cooker described in any one of the above.

[0177] In this embodiment, the trained temperature prediction model is deployed on the embedded processor or edge computing device of the induction cooker to achieve real-time temperature prediction.

[0178] In practical applications, the type and material of the cookware are set by the user according to the actual situation, the temperature at the center point of the cookware is collected in real time by a temperature sensor, and the power output gear of each plasma electrode is collected by the embedded processor. The target input vector including the type and material of the cookware, the temperature at the center point of the cookware at the current moment, and the working state information of the plasma electrodes of the induction cooker at the current moment is input into the temperature prediction model trained by the historical sample data set. The temperature prediction model can output the target predicted temperature at the next moment in real time. Finally, the embedded processor or edge computing device dynamically adjusts the power output gear of the plasma electrodes of the induction cooker according to the relationship between the target predicted temperature and the preset operating temperature, so as to realize the dynamic adjustment of the temperature of the cookware.

[0179] Fourthly, an embodiment of the present invention further provides a computer-readable storage medium, in which at least one instruction, at least one program, a code set or an instruction set is stored, and the at least one instruction, the at least one program, the code set or the instruction set is loaded and executed by a processor to implement the precise heating control method of the electric stove in the embodiment of the present invention.

[0180] In the above embodiment, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the processes or functions according to the embodiments of the present invention are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from a website, a computer, a server or a data center to another website, a computer, a server or a data center by wire (such as coaxial cable, optical fiber, digital subscriber line (DSL)) or wirelessly (such as infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or a data center that includes one or more integrated available media. The available medium can be a magnetic medium (for example, a floppy disk, a hard disk, a magnetic tape), an optical medium (for example, a DVD), or a semiconductor medium (for example, a solid state disk (SSD)).

[0181] As described above, the above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions within the technical scope disclosed by the present invention should be covered by the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

Claims

1. A precise heating control method for an electric stove, characterized in that Including: Construct a target input vector according to the obtained cookware type, cookware material, the temperature of the center point of the cookware at the current moment, and the working state information of the plasma electrodes of the induction cooker at the current moment; the working state information includes the power output gears of all the plasma electrodes of the induction cooker; Input the target input vector into the temperature prediction model trained by the historical sample data set, and predict the target prediction temperature at the next moment of the target input vector through the temperature prediction model; the historical sample data set includes historical input vectors and the actual temperatures at the next moment corresponding to the historical input vectors; Adjust the power output strategy of the plasma electrodes of the induction cooker according to the relationship between the target prediction temperature and the preset working temperature; When the target prediction temperature is less than the preset working temperature, adjust the power output strategy of the plasma electrodes of the induction cooker according to the relationship between the target prediction temperature and the preset working temperature, specifically including: Combined with the working state information of the plasma electrodes at the current moment, randomly turn on an unopened plasma electrode; And / or, randomly select half of the opened plasma electrodes, and increase the power output gear of the plasma electrodes whose power output gears have not reached the highest gear by one gear; When the target prediction temperature is greater than the preset working temperature, adjust the power output strategy of the plasma electrodes of the induction cooker according to the relationship between the target prediction temperature and the preset working temperature, specifically including: Combined with the working state information of the plasma electrodes at the current moment, randomly turn off an opened plasma electrode; And / or, randomly select half of the opened plasma electrodes, and reduce the power output gear of the plasma electrodes by one gear.

2. The precise heating control method of the electric stove according to claim 1, characterized in that Before predicting the target prediction temperature at the next moment of the target input vector through the temperature prediction model, the method further includes: Conduct a temperature measurement experiment on all combinations of the working state information of all the plasma electrodes of the induction cooker, the cookware material information, and the cookware type information, obtain the temperature of the center point of the cookware at the current moment for each combination, and the actual temperature of the center point of the cookware at the next moment; Construct a historical input vector for each combination according to the cookware type, cookware material, the temperature of the center point of the cookware at the current moment, and the working state information of the plasma electrodes at the current moment of each combination; Construct a historical sample data set according to multiple historical input vectors and the actual temperature of the center point of the cookware at the next moment corresponding to each historical input vector; Train the original temperature prediction model through the historical sample data set until the loss value between the predicted temperature output by the temperature prediction model and the actual temperature no longer decreases, and obtain the trained temperature prediction model.

3. The precise heating control method of the electric stove according to claim 2, wherein The loss value is calculated by the comprehensive loss function of the temperature prediction model; The comprehensive loss function is the sum of the temperature error loss and the energy consumption loss; For a sample group in the historical sample data set, the temperature error loss is the mean square error between the predicted temperature and the actual temperature, and the energy consumption loss is determined according to the power output gears of all the plasma electrodes of the induction cooker.

4. The precise heating control method for an electric stove according to claim 3, characterized in that, The loss function of the energy consumption loss is specifically: ； Among them, represents the energy consumption loss, represents the weight factor of the energy consumption loss in the comprehensive loss function, represents the number of the current sample in the sample group, represents the total number of samples in the sample group, represents the number of the plasma electrode in the current sample, represents the total number of plasma electrodes, represents the th power output gear of the th plasma electrode in the th sample.

5. The precise heating control method of the electric stove according to claim 1, characterized in that Before predicting the target predicted temperature of the target input vector at the next moment through the temperature prediction model, the method further includes: constructing an original temperature prediction model based on a multi-layer perceptron; The original temperature prediction model includes an input layer, a first hidden layer, a second hidden layer, and an output layer connected in sequence; The input layer is used to receive an input vector; the input vector is a historical input vector or a target input vector; The first hidden layer and the second hidden layer are sequentially arranged between the input layer and the output layer, and the second hidden layer is connected to the input layer with a residual connection; the first hidden layer and the second hidden layer are used to perform non-linear transformation and feature extraction on the received input vector; The output layer is used to receive the information of the second hidden layer and output a predicted temperature.

6. An accurate heating control device for an electric stove, characterized in that, Including: A data acquisition module, configured to construct a target input vector according to the obtained cookware type, cookware material, the temperature of the center point of the cookware at the current moment, and the working state information of the plasma electrodes of the induction cooker at the current moment; the working state information includes the power output gears of all the plasma electrodes of the induction cooker; A data processing module, configured to input the target input vector into a temperature prediction model trained by a historical sample data set, and predict the target predicted temperature of the target input vector at the next moment through the temperature prediction model; the historical sample data set includes a historical input vector and the actual temperature at the next moment corresponding to the historical input vector; A strategy adjustment module, configured to adjust the power output strategy of the plasma electrodes of the induction cooker according to the relationship between the target predicted temperature and a preset working temperature; When the target predicted temperature is less than the preset working temperature, the strategy adjustment module is configured to execute: Combined with the working state information of the plasma electrodes at the current moment, randomly turn on an unopened plasma electrode; And / or, randomly select half of the opened plasma electrodes, and increase the power output gear of the plasma electrodes whose power output gear has not reached the highest gear by one gear; When the target predicted temperature is greater than the preset working temperature, the strategy adjustment module is configured to execute: Combined with the working state information of the plasma electrodes at the current moment, randomly turn off an opened plasma electrode; And / or, randomly select half of the opened plasma electrodes, and reduce the power output gear of the plasma electrodes by one gear.

7. A device using plasma heating, characterized in that, Including: A controller; A memory for storing executable instructions of the controller; Wherein, the controller is configured to execute the instructions to implement the induction cooker precise heating control method according to any one of claims 1-5.

8. A computer-readable storage medium, characterized in that, At least one instruction, at least one program, a code set or an instruction set is stored in the storage medium, and the at least one instruction, the at least one program, the code set or the instruction set is loaded and executed by a processor to implement the induction cooker precise heating control method according to any one of claims 1-5.

Images