Method and system for testing energy efficiency of air fryer

Abstract

The invention discloses a method and a system for testing energy efficiency of an air fryer. The method comprises the following steps: acquiring load data, temperature monitoring data and power consumption of the air fryer; obtaining energy efficiency data based on the load data and the power consumption, and performing clustering analysis on the temperature monitoring data in the preheating stage and the cooling stage to obtain a thermal inertia index; performing stability analysis on the temperature monitoring data to obtain a temperature control deviation rate, and predicting the energy efficiency data according to the temperature control deviation rate and the thermal inertia index to obtain a predicted energy efficiency sequence; constructing a comprehensive energy efficiency index based on the predicted energy efficiency sequence, and establishing an air fryer energy efficiency test model by adopting the comprehensive energy efficiency index; and optimizing the air fryer energy efficiency test model through a dogvessel colony algorithm, inputting to-be-tested data into the air fryer energy efficiency test model, and outputting a test result. According to the method, through accurate calculation and comprehensive analysis, cognition on the energy efficiency of the air fryer is remarkably improved, and the method plays a role in promoting industry development.

Description

Technical Field

[0001] The present invention relates to the technical field of energy efficiency testing of household appliances, and in particular to an energy efficiency testing method and system for an air fryer. Background Art

[0002] As people pay more and more attention to healthy eating, air fryers, as a kitchen appliance that can cook food in a relatively healthy way, have been favored by more and more consumers. However, there are large differences in energy efficiency among air fryers on the market, which not only affects the user's usage cost, but also relates to the rational use of energy and environmental protection.

[0003] At present, the test methods for the energy efficiency of air fryers are still imperfect. Traditional testing methods often focus on a single factor, such as power consumption or temperature control accuracy, while ignoring the impact of other important factors on energy efficiency. For example, during the cooking process, the characteristics of the load (such as food), the energy loss in the preheating and cooling stages, and the stability of the temperature will have a significant impact on the energy efficiency of the air fryer. At the same time, the lack of a systematic testing method and model that comprehensively considers multiple factors makes it difficult to accurately evaluate the energy efficiency performance of the air fryer in actual use.

[0004] Therefore, it is necessary to develop a more comprehensive and accurate air fryer energy efficiency testing method and system to meet market demand and promote the sustainable development of the air fryer industry. Summary of the invention

[0005] The purpose of the present invention is to provide an air fryer energy efficiency testing method and system.

[0006] To achieve the above object, the present invention is implemented according to the following technical solutions:

[0007] The present invention comprises the following steps:

[0008] Get the load data, temperature monitoring data and power consumption of the air fryer;

[0009] Energy efficiency data is obtained based on the load data and the power consumption, and a thermal inertia index is obtained by clustering the temperature monitoring data in the preheating stage and the cooling stage;

[0010] Performing stability analysis on the temperature monitoring data to obtain a temperature control deviation rate, and predicting the energy efficiency data according to the temperature control deviation rate and the thermal inertia index to obtain a predicted energy efficiency sequence;

[0011] Constructing a comprehensive energy efficiency index based on the predicted energy efficiency sequence, and using the comprehensive energy efficiency index to establish an air fryer energy efficiency test model;

[0012] The air fryer energy efficiency test model is optimized by the salp swarm algorithm, and the data to be tested is input into the air fryer energy efficiency test model to output the test result.

[0013] Furthermore, the method for obtaining the load data, temperature monitoring data and power consumption of the air fryer includes:

[0014] Frozen French fries or frozen chicken wings are used as loads. The load data includes the mass data of the load before and after baking. A thermocouple is used to measure the temperature in the center of the air fryer cavity to obtain temperature monitoring data. The three time periods of the temperature monitoring data include the preheating stage, the baking load stage, and the cooling stage of the air fryer after power failure. A high-precision electric energy meter is used to measure the power consumption of the air fryer during operation.

[0015] Furthermore, the method of obtaining energy efficiency data based on the load data and the power consumption and obtaining a thermal inertia index by clustering analysis of the temperature monitoring data in the preheating stage and the cooling stage includes:

[0016] Energy efficiency data are:

[0017]

[0018] Where φ(t) is the energy efficiency data, E is the total power consumption, m1 is the initial mass of the frozen French fries, m2 is the mass of the frozen French fries after baking, B is the time attenuation coefficient, and t is the baking time;

[0019] The preheating time is obtained based on the temperature monitoring data in the preheating stage. The preheating time is the time taken by the air fryer to heat up to the set temperature. The cooling time is obtained based on the temperature monitoring data in the cooling stage. The cooling time is the time taken by the air fryer to cool down to the reference temperature.

[0020] The preheating time, cooling time and power consumption in the preheating stage are normalized and preprocessed to form data points, and the clustering model is trained:

[0021] Estimation of kernel density:

[0022]

[0023] in For The kernel density estimated using the kernel function K and kernel width h, K(·) is the kernel function K, using the Gaussian kernel function, ψ(·) is the wavelet function, is the jth cluster S j The center point, c K is a constant, N is the total number of data points, p is the dimension of the data points, and w i is the weight of the i-th data point, X i is the i-th data point, ‖·‖ represents the Euclidean distance,

[0024] Gradient of the kernel density estimate:

[0025]

[0026] in for The gradient of, g is a function defined based on the kernel function K and the wavelet function, g(X) = -ψ ′ (X)K ′ (X), K ′ (·) is the derivative of the kernel function K, ψ ′ (·) is the derivative of ψ(·),

[0027] The formula for the mean shift vector is:

[0028]

[0029] in is the mean shift vector, c g is a constant coefficient used to adjust the scaling of g, For The kernel density estimated using the kernel function G and kernel width h, G(X) = c g g(‖X‖ 2 ), iteratively calculate the mean shift vector and update the center point of the cluster.

[0030] Calculate the average power consumption of the preheating time, cooling time and preheating stage in the cluster respectively to get the thermal inertia index of the data point:

[0031]

[0032] in, and are the weight coefficients of the preheating time, cooling time and average power consumption in the preheating stage in the cluster where the data point is located. and are the average values ​​of the preheating time, cooling time and power consumption in the preheating stage in the cluster where the data point is located.

[0033] Furthermore, a method for performing stability analysis on the temperature monitoring data to obtain a temperature control deviation rate includes:

[0034] The temperature change rate, deviation value relative to the set temperature and standard deviation of the temperature monitoring data in the baking load stage are extracted, and the temperature control deviation rate is obtained by weighted summation after normalization.

[0035] Furthermore, the method of predicting the energy efficiency data according to the temperature control deviation rate and the thermal inertia index to obtain a predicted energy efficiency sequence includes:

[0036] The temperature control deviation rate, load change mass, set temperature and thermal inertia index are used as characteristic data, and energy efficiency data is used as the target value to establish a prediction model. The objective function is:

[0037]

[0038] Where L(θ) is the objective function, n is the number of target values, and y i is the i-th target value, is the pair y obtained in the first t-1 iterations i The predicted value, l(·) is the loss function, f t (x i ) is the decision tree added in the tth iteration at x i The value of x i is the i-th feature data, K is the number of decision trees, f k is the kth decision tree, Ω(·) is the regularization term used to control the complexity of the prediction model, β is the depth weight, and D is the maximum depth of the leaf nodes in the decision tree;

[0039] The objective function after the second-order Taylor expansion is:

[0040]

[0041]

[0042] in, is the loss function about The first derivative of is the loss function about The second-order derivative of f t is the decision tree added in the tth iteration, γ and λ are regularization parameters, T is the number of leaf nodes, and w j is the weight of the jth leaf node;

[0043] When the objective function is about w j When the derivative of is 0, the optimal weight of the leaf node is obtained:

[0044]

[0045] in is the optimal weight of leaf node j, d j is the depth of leaf node j, I j ={i|q(x i )=j}, is the sample set of the jth leaf node, where q(xi ) represents x i The leaf node number to which it belongs, Substitute it into the objective function and get the optimal value of the objective function:

[0046]

[0047] For each feature data, divide its value range into equal intervals and count the number of samples falling in each interval.

[0048] The calculation formula for split gain is:

[0049]

[0050] Among them, I is the sample set of the current node, I L and I R are the sample sets of the left child node and the right child node after splitting,

[0051] Select the feature data and interval with the largest splitting gain as the splitting node;

[0052] The quality changes of different loads are input into the trained prediction model, and the prediction sequence of energy efficiency data is output.

[0053] Furthermore, a method for constructing a comprehensive energy efficiency index based on the predicted energy efficiency sequence includes:

[0054] Calculate the average value of the forecast sequence and construct a comprehensive energy efficiency index:

[0055]

[0056] Where Z v is the comprehensive energy efficiency index, z q is the temperature control deviation rate, z p is the average value of the prediction sequence, α is the environmental temperature influence factor, T env is the ambient temperature data, T d To set the temperature.

[0057] Furthermore, a method for establishing an energy efficiency test model for an air fryer using a comprehensive energy efficiency index includes:

[0058] The temperature control deviation rate and thermal inertia index are used as independent variables, and the comprehensive energy efficiency index is used as the dependent variable to establish a data set. The data set is divided into a training set and a test set in a ratio of 3:1. The neural network algorithm is used to learn the mapping relationship between the independent variables and the dependent variables through the training set, and the air fryer energy efficiency test model is established. The test set is used to optimize the hyperparameters of the constant temperature energy efficiency test model.

[0059] Furthermore, the method for optimizing the air fryer energy efficiency test model by using a salp swarm algorithm comprises:

[0060] The salp chain consists of two types of salps: leader and follower. The leader is at the front of the salp chain, and the other salps are followers. The salp population vector consists of N salps in d dimensions, where d is the number of hyperparameters. The position of each salp represents a set of hyperparameter values. The population is initialized using the Tent mapping formula:

[0061]

[0062] in, and Respectively represent the number of salps in the Second and The chaotic variable value after the secondary chaotic mapping, μ∈(0,2], is the chaotic parameter, i represents the index of the salp in the population, Represents the chaotic variable number;

[0063] The chaotic sequence after Tent chaos mapping is inversely mapped to the corresponding hyperparameter search space to obtain the position sequence, and the fitness value of each salp is calculated. The optimal individual is used as the food source position. The fitness value is the inverse of the mean square error of the constant temperature energy efficiency test model.

[0064] Leader position update formula:

[0065]

[0066] in, is the position of the first salp as the leader in the j-dimensional space, F j is the location of the food source in the j-dimensional space, c4 is a random number uniformly distributed between [0, 1], is a very small constant, take 0.0001, P cr For the preset crazy probability, ub j and lb j are the upper and lower limits of the j-th dimension space, c2 and c3 are random numbers generated in the interval [0, 1]. Where e is the base of the natural logarithm, t is the current number of iterations, T max is the maximum number of iterations;

[0067] Follower position update formula:

[0068]

[0069] in, is the position of the i-th salp in the j-dimensional space, ω(t) is the adaptive inertia weight, ω s is the initial inertia weight, ω e is the inertia weight when the iteration reaches the maximum number of times;

[0070] After each iteration, the fitness value of each salp is calculated, and the optimal individual is used as the food source position. The iterative update is performed until the maximum number of iterations is reached, and the position corresponding to the maximum fitness value is used as the optimal hyperparameter value.

[0071] An air fryer energy efficiency testing system, characterized by comprising:

[0072] The acquisition module obtains the load data, temperature monitoring data and power consumption of the air fryer;

[0073] a calculation module, which obtains energy efficiency data based on the load data and the power consumption, performs cluster analysis on the temperature monitoring data in the preheating stage and the cooling stage to obtain a thermal inertia index, performs stability analysis on the temperature monitoring data to obtain a temperature control deviation rate, predicts the energy efficiency data according to the temperature control deviation rate and the thermal inertia index to obtain a predicted energy efficiency sequence, constructs a comprehensive energy efficiency index based on the predicted energy efficiency sequence, and uses the comprehensive energy efficiency index to establish an air fryer energy efficiency test model;

[0074] An optimization module, which optimizes the air fryer energy efficiency test model by using a salp swarm algorithm;

[0075] The output module inputs the test data into the air fryer energy efficiency test model to output the test results.

[0076] The beneficial effects of the present invention are:

[0077] The present invention comprehensively considers the impact of multiple factors on the energy efficiency of the air fryer, including load data, temperature monitoring data, power consumption, etc. By analyzing and processing these data, the energy efficiency performance of the air fryer can be evaluated more comprehensively and accurately, avoiding the limitation of traditional testing methods that only focus on a single factor. BRIEF DESCRIPTION OF THE DRAWINGS

[0078] Figure 1 The present invention is a flow chart of the steps of an air fryer energy efficiency testing method and system thereof. DETAILED DESCRIPTION

[0079] The present invention is further described below by means of specific embodiments. The illustrative embodiments and descriptions of the present invention are used to explain the present invention but are not intended to limit the present invention.

[0080] The present invention provides an air fryer energy efficiency testing method and system thereof, comprising the following steps:

[0081] like Figure 1 As shown, in this embodiment, the following steps are included:

[0082] Get the load data, temperature monitoring data and power consumption of the air fryer;

[0083] In the actual evaluation, the relative humidity was 55% and the atmospheric pressure was 101.325 kPa. Three air fryers with a capacity of 3 liters and a rated power of 1000 W were selected for testing. Frozen French fries were selected as the load. After the three air fryers were preheated to the set temperature of 180 degrees, the load was added for baking. The color card was used for comparison. The load was baked until the color was uniformly light yellow and then the power was turned off. The mass of the load before baking was uniformly 300 g. The mass of the load after baking was: 210 g for No. 1 air fryer, 205 g for No. 2 air fryer, and 300 g for No. 3 air fryer. Air fryer 200g; temperature monitoring data showed that the preheating time for No. 1, 2 and 3 were [3 minutes, 2.5 minutes, 3.5 minutes], the baking time was [5 minutes, 4.5 minutes, 5.5 minutes], and the cooling time was [15 minutes, 13.5 minutes, 16.5 minutes]; the preheating power consumption was [0.05KW / h, 0.04KW / h, 0.06KW / h]: the total power consumption was [0.35KW / h, 0.32KW / h, 0.38KW / h];

[0084] Energy efficiency data is obtained based on the load data and the power consumption, and a thermal inertia index is obtained by clustering the temperature monitoring data in the preheating stage and the cooling stage;

[0085] In the actual evaluation, the energy efficiency data were [3.24, 2.86, 3.66], the temperature monitoring data of No. 1 and No. 2 were in one category, No. 3 was in a separate category, and the thermal inertia index was [0.034, 0.034, 0.04];

[0086] Performing stability analysis on the temperature monitoring data to obtain a temperature control deviation rate, and predicting the energy efficiency data according to the temperature control deviation rate and the thermal inertia index to obtain a predicted energy efficiency sequence;

[0087] In the actual evaluation, the temperature control deviation rates are [0.2, 0.19, 0.21] respectively; the predicted energy efficiency sequences for load changes of 80g, 110g, and 120g are: No. 1 [2.81, 4.04, 4.42], No. 2 [2.21, 3.24, 4.02], and No. 3 [2.82, 4.05, 4.43];

[0088] Constructing a comprehensive energy efficiency index based on the predicted energy efficiency sequence, and using the comprehensive energy efficiency index to establish an air fryer energy efficiency test model;

[0089] In the actual evaluation, the ambient temperature impact factor is 0.0005, and the comprehensive energy efficiency indexes are: [3.26, 2.78, 3.23];

[0090] The air fryer energy efficiency test model is optimized by the salp swarm algorithm, and the data to be tested is input into the air fryer energy efficiency test model to output the test result.

[0091] In this embodiment, the method for obtaining load data, temperature monitoring data and power consumption of the air fryer includes:

[0092] Frozen French fries or frozen chicken wings are used as loads. The load data includes the mass data of the load before and after baking. A thermocouple is used to measure the temperature in the center of the air fryer cavity to obtain temperature monitoring data. The three time periods of the temperature monitoring data include the preheating stage, the baking load stage, and the cooling stage of the air fryer after power failure. A high-precision electric energy meter is used to measure the power consumption of the air fryer during operation.

[0093] In this embodiment, the method of obtaining energy efficiency data based on the load data and the power consumption, and obtaining a thermal inertia index by clustering analysis of the temperature monitoring data in the preheating stage and the cooling stage includes:

[0094] Energy efficiency data are:

[0095]

[0096] Where φ(t) is the energy efficiency data, E is the total power consumption, m1 is the initial mass of the frozen French fries, m2 is the mass of the frozen French fries after baking, B is the time attenuation coefficient, and t is the baking time;

[0097] The preheating time is obtained according to the temperature monitoring data in the preheating stage. The preheating time is the time taken by the air fryer to heat up to the set temperature. The cooling time is obtained according to the temperature monitoring data in the cooling stage. The cooling time is the time taken by the air fryer to cool down to the reference temperature. The reference temperature is 100°C.

[0098] The preheating time, cooling time and power consumption in the preheating stage are normalized and preprocessed to form data points, and the clustering model is trained:

[0099] Estimation of kernel density:

[0100]

[0101] in For The kernel density estimated using the kernel function K and kernel width h, K(·) is the kernel function K, using the Gaussian kernel function, ψ(·) is the wavelet function, is the jth cluster S j The center point, cK is a constant, N is the total number of data points, p is the dimension of the data points, and w i is the weight of the i-th data point, X i is the i-th data point, ‖·‖ represents the Euclidean distance,

[0102] Gradient of the kernel density estimate:

[0103]

[0104] in for The gradient of g is the function defined based on the kernel function K and the wavelet function, g(X) = -ψ ′ (X)K ′ (X), K ′ (·) is the derivative of the kernel function K, ψ ′ (·) is the derivative of ψ(·),

[0105] The formula for the mean shift vector is:

[0106]

[0107] in is the mean shift vector, c g is a constant coefficient used to adjust the scaling of g, For The kernel density estimated using the kernel function G and kernel width h, G(X) = c g g(‖X‖ 2 ), iteratively calculate the mean shift vector and update the center point of the cluster.

[0108] Calculate the average power consumption of the preheating time, cooling time and preheating stage in the cluster respectively to get the thermal inertia index of the data point:

[0109]

[0110] Among them, r y 、r yh and r j are the weight coefficients of the preheating time, cooling time and average power consumption in the preheating stage in the cluster where the data point is located. and are the average values ​​of the preheating time, cooling time and power consumption in the preheating stage in the cluster where the data point is located.

[0111] In this embodiment, the method of performing stability analysis on the temperature monitoring data to obtain the temperature control deviation rate includes:

[0112] The temperature change rate, deviation value relative to the set temperature and standard deviation of the temperature monitoring data in the baking load stage are extracted, and the temperature control deviation rate is obtained by weighted summation after normalization.

[0113] In this embodiment, the method of predicting the energy efficiency data according to the temperature control deviation rate and the thermal inertia index to obtain a predicted energy efficiency sequence includes:

[0114] The temperature control deviation rate, load change mass, set temperature and thermal inertia index are used as characteristic data, and energy efficiency data is used as the target value to establish a prediction model. The objective function is:

[0115]

[0116] Where L(θ) is the objective function, n is the number of target values, and y i is the i-th target value, is the pair y obtained in the first t-1 iterations i The predicted value, l(·) is the loss function, f t (x i ) is the decision tree added in the tth iteration at x i The value of x i is the i-th feature data, K is the number of decision trees, f k is the kth decision tree, Ω(·) is the regularization term used to control the complexity of the prediction model, β is the depth weight, and D is the maximum depth of the leaf nodes in the decision tree;

[0117] The objective function after the second-order Taylor expansion is:

[0118]

[0119] in, is the loss function about The first derivative of is the loss function about The second-order derivative of f t is the decision tree added in the tth iteration, γ and λ are regularization parameters, T is the number of leaf nodes, and w j is the weight of the jth leaf node;

[0120] When the objective function is about w j When the derivative of is 0, the optimal weight of the leaf node is obtained:

[0121]

[0122] in is the optimal weight of leaf node j, d j is the depth of leaf node j, I j ={i|q(x i )=j}, is the sample set of the jth leaf node, where q(x i ) represents x i The leaf node number to which it belongs, Substitute it into the objective function and get the optimal value of the objective function:

[0123]

[0124] For each feature data, divide its value range into equal intervals and count the number of samples falling in each interval.

[0125] The calculation formula for split gain is:

[0126]

[0127] Among them, I is the sample set of the current node, I L and I R are the sample sets of the left child node and the right child node after splitting,

[0128] Select the feature data and interval with the largest splitting gain as the splitting node;

[0129] The quality changes of different loads are input into the trained prediction model, and the prediction sequence of energy efficiency data is output.

[0130] In this embodiment, the method for constructing a comprehensive energy efficiency index based on the predicted energy efficiency sequence includes:

[0131] Calculate the average value of the forecast sequence and construct a comprehensive energy efficiency index:

[0132]

[0133] Where Z v is the comprehensive energy efficiency index, z q is the temperature control deviation rate, z p is the average value of the prediction sequence, α is the environmental temperature influence factor, T env is the ambient temperature data, T d To set the temperature.

[0134] In this embodiment, the method of establishing an air fryer energy efficiency test model using a comprehensive energy efficiency index includes:

[0135] The temperature control deviation rate and thermal inertia index are used as independent variables, and the comprehensive energy efficiency index is used as the dependent variable to establish a data set. The data set is divided into a training set and a test set in a ratio of 3:1. The neural network algorithm is used to learn the mapping relationship between the independent variables and the dependent variables through the training set, and the air fryer energy efficiency test model is established. The test set is used to optimize the hyperparameters of the constant temperature energy efficiency test model.

[0136] In this embodiment, the method for optimizing the air fryer energy efficiency test model by using the salp swarm algorithm includes:

[0137] The salp chain consists of two types of salps: leader and follower. The leader is at the front of the salp chain, and the other salps are followers. The salp population vector consists of N salps in d dimensions, where d is the number of hyperparameters. The position of each salp represents a set of hyperparameter values. The population is initialized using the Tent mapping formula:

[0138]

[0139] in, and Respectively represent the number of salps in the Second and The chaotic variable value after the secondary chaotic mapping, μ∈(0,2], is the chaotic parameter, i represents the index of the salp in the population, Represents the chaotic variable number;

[0140] The chaotic sequence after Tent chaos mapping is inversely mapped to the corresponding hyperparameter search space to obtain the position sequence, and the fitness value of each salp is calculated. The optimal individual is used as the food source position. The fitness value is the inverse of the mean square error of the constant temperature energy efficiency test model.

[0141] Leader position update formula:

[0142]

[0143] in, is the position of the first salp as the leader in the j-dimensional space, F j is the location of the food source in the j-dimensional space, c4 is a random number uniformly distributed between [0, 1], is a very small constant, take 0.0001, P cr For the preset crazy probability, ub j and lb j are the upper and lower limits of the j-th dimension space, c2 and c3 are random numbers generated in the interval [0, 1]. Where e is the base of the natural logarithm, t is the current number of iterations, T maxis the maximum number of iterations;

[0144] Follower position update formula:

[0145]

[0146] in, is the position of the i-th salp in the j-dimensional space, ω(t) is the adaptive inertia weight, ω s is the initial inertia weight, ω e is the inertia weight when the iteration reaches the maximum number of times;

[0147] After each iteration, the fitness value of each salp is calculated, and the optimal individual is used as the food source position. The iterative update is performed until the maximum number of iterations is reached, and the position corresponding to the maximum fitness value is used as the optimal hyperparameter value.

[0148] An air fryer energy efficiency testing system, characterized by comprising:

[0149] The acquisition module obtains the load data, temperature monitoring data and power consumption of the air fryer;

[0150] a calculation module, which obtains energy efficiency data based on the load data and the power consumption, performs cluster analysis on the temperature monitoring data in the preheating stage and the cooling stage to obtain a thermal inertia index, performs stability analysis on the temperature monitoring data to obtain a temperature control deviation rate, predicts the energy efficiency data according to the temperature control deviation rate and the thermal inertia index to obtain a predicted energy efficiency sequence, constructs a comprehensive energy efficiency index based on the predicted energy efficiency sequence, and uses the comprehensive energy efficiency index to establish an air fryer energy efficiency test model;

[0151] An optimization module, which optimizes the air fryer energy efficiency test model by using a salp swarm algorithm;

[0152] The output module inputs the test data into the air fryer energy efficiency test model to output the test results.

[0153] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for testing the energy efficiency of an air fryer, characterized in that: The following steps are involved: Get the load data, temperature monitoring data and power consumption of the air fryer; Energy efficiency data is obtained based on the load data and the power consumption, and a thermal inertia index is obtained by clustering the temperature monitoring data in the preheating stage and the cooling stage; Performing stability analysis on the temperature monitoring data to obtain a temperature control deviation rate, and predicting the energy efficiency data according to the temperature control deviation rate and the thermal inertia index to obtain a predicted energy efficiency sequence; Constructing a comprehensive energy efficiency index based on the predicted energy efficiency sequence, and using the comprehensive energy efficiency index to establish an air fryer energy efficiency test model; The air fryer energy efficiency test model is optimized by the salp swarm algorithm, and the data to be tested is input into the air fryer energy efficiency test model to output the test result.

2. The air fryer energy efficiency testing method according to claim 1, characterized in that: The method for obtaining the load data, temperature monitoring data and power consumption of the air fryer includes: Frozen French fries or frozen chicken wings are used as loads. The load data includes the mass data of the load before and after baking. A thermocouple is used to measure the temperature in the center of the air fryer cavity to obtain temperature monitoring data. The three time periods of the temperature monitoring data include the preheating stage, the baking load stage, and the cooling stage of the air fryer after power failure. A high-precision electric energy meter is used to measure the power consumption of the air fryer during operation.

3. The air fryer energy efficiency testing method according to claim 1, characterized in that: The method of obtaining energy efficiency data based on the load data and the power consumption, and performing cluster analysis on the temperature monitoring data in the preheating stage and the cooling stage to obtain a thermal inertia index includes: Energy efficiency data are: Where φ(t) is the energy efficiency data, E is the total power consumption, m1 is the initial mass of the frozen French fries, m2 is the mass of the frozen French fries after baking, B is the time attenuation coefficient, and t is the baking time; The preheating time is obtained based on the temperature monitoring data in the preheating stage. The preheating time is the time taken by the air fryer to heat up to the set temperature. The cooling time is obtained based on the temperature monitoring data in the cooling stage. The cooling time is the time taken by the air fryer to cool down to the reference temperature. The preheating time, cooling time and power consumption in the preheating stage are normalized and preprocessed to form data points, and the clustering model is trained: Estimation of kernel density: in For The kernel density estimated using the kernel function K and kernel width h, K(·) is the kernel function K, using the Gaussian kernel function, ψ(·) is the wavelet function, is the jth cluster S j The center point, c K is a constant, N is the total number of data points, p is the dimension of the data points, and w i is the weight of the i-th data point, X i is the i-th data point, ‖·‖ represents the Euclidean distance, Gradient of the kernel density estimate: in for The gradient of , g is a function defined based on the kernel function K and the wavelet function, g(X) = -ψ′(X)K′(X), K′(·) is the derivative of the kernel function K, ψ′(·) is the derivative of ψ(·), The formula for the mean shift vector is: in is the mean shift vector, c g is a constant coefficient used to adjust the scaling of g, For The kernel density estimated using the kernel function G and kernel width h, G(X) = c g g(‖X‖ 2 ), iteratively calculate the mean shift vector and update the center point of the cluster. Calculate the average power consumption of the preheating time, cooling time and preheating stage in the cluster respectively to get the thermal inertia index of the data point: in, and are the weight coefficients of the preheating time, cooling time and average power consumption in the preheating stage in the cluster where the data point is located. and are the average values ​​of the preheating time, cooling time and power consumption in the preheating stage in the cluster where the data point is located.

4. The air fryer energy efficiency testing method according to claim 1, characterized in that: The method of performing stability analysis on the temperature monitoring data to obtain a temperature control deviation rate includes: The temperature change rate, deviation value relative to the set temperature and standard deviation of the temperature monitoring data in the baking load stage are extracted, and the temperature control deviation rate is obtained by weighted summation after normalization.

5. The air fryer energy efficiency testing method according to claim 1, characterized in that: The method of predicting the energy efficiency data according to the temperature control deviation rate and the thermal inertia index to obtain a predicted energy efficiency sequence includes: The temperature control deviation rate, load change mass, set temperature and thermal inertia index are used as characteristic data, and energy efficiency data is used as the target value to establish a prediction model. The objective function is: Where L(θ) is the objective function, n is the number of target values, and y i is the i-th target value, is the pair y obtained in the first t-1 iterations i The predicted value, l(·) is the loss function, f t (x i ) is the decision tree added in the tth iteration at x i The value of x i is the i-th feature data, K is the number of decision trees, f k is the kth decision tree, Ω(·) is the regularization term used to control the complexity of the prediction model, β is the depth weight, and D is the maximum depth of the leaf nodes in the decision tree; The objective function after the second-order Taylor expansion is: in, is the loss function about The first derivative of is the loss function about The second-order derivative of f t is the decision tree added in the tth iteration, γ and λ are regularization parameters, T is the number of leaf nodes, and w j is the weight of the jth leaf node; When the objective function is about w j When the derivative of is 0, the optimal weight of the leaf node is obtained: in is the optimal weight of leaf node j, d j is the depth of leaf node j, I j ={i|q(x i )=j}, is the sample set of the jth leaf node, where q(x i ) represents x i The leaf node number to which it belongs, Substitute it into the objective function and get the optimal value of the objective function: For each feature data, divide its value range into equal intervals and count the number of samples falling in each interval. The calculation formula for split gain is: Among them, I is the sample set of the current node, I L and I R are the sample sets of the left child node and the right child node after splitting, Select the feature data and interval with the largest splitting gain as the splitting node; The mass changes of different loads are input into the trained prediction model, and the prediction sequence of energy efficiency data is output.

6. The air fryer energy efficiency testing method according to claim 1, characterized in that: The method for constructing a comprehensive energy efficiency index based on the predicted energy efficiency sequence includes: Calculate the average value of the forecast sequence and construct a comprehensive energy efficiency index: Where Z v is the comprehensive energy efficiency index, is the temperature control deviation rate, is the average value of the prediction sequence, α is the environmental temperature influence factor, T env is the ambient temperature data, T d To set the temperature.

7. The air fryer energy efficiency testing method according to claim 1, characterized in that: The method of establishing an air fryer energy efficiency test model using a comprehensive energy efficiency index includes: The temperature control deviation rate and thermal inertia index are used as independent variables, and the comprehensive energy efficiency index is used as the dependent variable to establish a data set. The data set is divided into a training set and a test set in a ratio of 3:

1. The neural network algorithm is used to learn the mapping relationship between the independent variables and the dependent variables through the training set, and the air fryer energy efficiency test model is established. The test set is used to optimize the hyperparameters of the constant temperature energy efficiency test model.

8. The air fryer energy efficiency testing method according to claim 1, characterized in that: The method for optimizing the air fryer energy efficiency test model by using a salp swarm algorithm comprises: The salp chain consists of two types of salps: leader and follower. The leader is at the front of the salp chain, and the other salps are followers. The salp population vector consists of N salps in d dimensions, where d is the number of hyperparameters. The position of each salp represents a set of hyperparameter values. The population is initialized using the Tent mapping formula: in, and Respectively represent the number of salps in the Second and The chaotic variable value after the secondary chaotic mapping, μ∈(0,2], is the chaotic parameter, i represents the index of the salp in the population, Represents the chaotic variable number; The chaotic sequence after Tent chaos mapping is inversely mapped to the corresponding hyperparameter search space to obtain the position sequence, and the fitness value of each salp is calculated. The optimal individual is used as the food source position. The fitness value is the inverse of the mean square error of the constant temperature energy efficiency test model. Leader position update formula: in, is the position of the first salp as the leader in the j-dimensional space, F j is the location of the food source in the j-dimensional space, c4 is a random number uniformly distributed between [0, 1], is a very small constant, take 0.0001, P cr For the preset crazy probability, ub j and lb j are the upper and lower limits of the j-th dimension space, c2 and c3 are random numbers generated in the interval [0, 1]. Where e is the base of the natural logarithm, t is the current number of iterations, T max is the maximum number of iterations; Follower position update formula: in, is the position of the i-th salp in the j-dimensional space, ω(t) is the adaptive inertia weight, ω s is the initial inertia weight, ω e is the inertia weight when the iteration reaches the maximum number of times; After each iteration, the fitness value of each salp is calculated, and the optimal individual is used as the food source position. The iterative update is performed until the maximum number of iterations is reached, and the position corresponding to the maximum fitness value is used as the optimal hyperparameter value.

9. An air fryer energy efficiency testing system, used to execute the air fryer energy efficiency testing method according to any one of claims 1 to 8, characterized in that: include: The acquisition module obtains the load data, temperature monitoring data and power consumption of the air fryer; a calculation module, which obtains energy efficiency data based on the load data and the power consumption, performs cluster analysis on the temperature monitoring data in the preheating stage and the cooling stage to obtain a thermal inertia index, performs stability analysis on the temperature monitoring data to obtain a temperature control deviation rate, predicts the energy efficiency data according to the temperature control deviation rate and the thermal inertia index to obtain a predicted energy efficiency sequence, constructs a comprehensive energy efficiency index based on the predicted energy efficiency sequence, and uses the comprehensive energy efficiency index to establish an air fryer energy efficiency test model; An optimization module, which optimizes the air fryer energy efficiency test model by using a salp swarm algorithm; The output module inputs the test data into the air fryer energy efficiency test model to output the test results.

Images