Energy regulation method for zero-carbon operation of wind-solar-storage building considering personnel thermal comfort

Abstract

The wind-solar-storage building zero-carbon operation energy regulation method considering personnel thermal comfort degree disclosed by the present application belongs to the field of carbon emission reduction in the operation stage of small buildings. The implementation method of the present application is as follows: the energy supply and consumption boundary of the small building is established, the personnel activity sample data set and the wind-solar natural endowment data set are obtained; the principal components of human activity data are analyzed and the personnel energy use behavior is determined based on the PCA-DA comprehensive method; the dimensionality-reduced activity data is subjected to cluster analysis and the energy use behavior is determined by using the discriminant analysis method DA; the dynamic threshold of the energy consumption of the existing electrical appliances of the small building meeting the wind-solar natural endowment is dynamically obtained by using a deep learning model; the human comfort comprehensive evaluation index and the comfort function are constructed; based on the obtained dynamic threshold range of the energy consumption of the existing electrical appliances of the small building, the energy use priority of the electrical appliances of the small building is dynamically adjusted, the energy regulation strategy is selected and set in combination with the comfort evaluation result, and the wind-solar-storage small building autonomous energy supply and zero-carbon operation duration is prolonged.

Description

Technical Field

[0001] This invention relates to an energy regulation method for zero-carbon operation of small wind-solar-storage buildings that takes into account the thermal comfort of personnel, and belongs to the field of carbon emission reduction during the operation of small buildings. Background Technology

[0002] Among the many low-carbon building technologies, wind-solar-storage integrated buildings are a crucial category because they can supply clean energy using wind turbines and solar photovoltaic panels, and store energy through an electrical energy storage system. They also have the hardware foundation for energy regulation, making it possible for the entire building to operate with zero carbon emissions.

[0003] Due to the inherent uncertainties of natural phenomena such as sunlight and wind, the energy supply of small-scale buildings relying primarily on wind and solar power generation is unstable. Based on an effective estimate of energy supply, achieving reasonable control over electricity consumption is crucial to reducing the use of energy storage components with high carbon emissions and truly realizing long-term zero-carbon operation of buildings.

[0004] In addition, compared with conventional medium and large-sized buildings, small buildings have a larger ratio of total building envelope area to usable area, a worse thermal environment, and are more difficult to manage for thermal comfort of personnel. Moreover, their structural composition is relatively simple, with only three types of energy hardware: wind, solar and energy storage. They lack subsystems such as hot water storage tanks and photovoltaic wall panels that can be used for energy regulation, making energy regulation more difficult and thus affecting the achievement of the zero-carbon operation goal of wind, solar and energy storage small buildings. Summary of the Invention

[0005] The purpose of this invention is to provide an energy regulation method for zero-carbon operation of wind, solar and energy storage buildings that takes into account the thermal comfort of people. This method can ensure that the energy supply system of small zero-carbon buildings uses wind, solar and energy storage systems to supply energy, while ensuring human comfort and users' energy needs in the long term, and maintaining the autonomous operation of small zero-carbon buildings.

[0006] The objective of this invention is achieved through the following technical solution:

[0007] This invention discloses a zero-carbon operation energy regulation method for wind-solar-storage buildings that considers human thermal comfort. It establishes energy supply and consumption boundaries for small buildings equipped with independent microgrids. Smart sensors are installed inside the building at energy-consuming locations, on functional equipment, and on the building envelope to acquire a dataset of human activity patterns and a dataset of wind and solar natural endowments. This includes the on / off times of existing indoor appliances, their usage duration, energy consumption, usage duration of appliances carried by personnel, energy consumption when these appliances are connected to the building's energy supply, the time of personnel entering and leaving the building, changes made to the building envelope by personnel (opening windows, doors), and energy losses in the building's energy storage system. The method uses a PCA-DA synthesis method to analyze the principal components of the human activity data and determine... For personnel energy consumption behavior, discriminant analysis (DA) is used to cluster the dimensionality-reduced activity data and determine energy consumption behavior. Based on the clustered data, a deep learning model is used to dynamically obtain the dynamic thresholds for energy consumption of existing electrical appliances in small buildings that meet the natural endowments of wind and solar power. The system accesses meteorological forecast data and, combined with the dynamic thresholds for energy consumption of electrical appliances obtained in the previous step, sets the optimal battery storage configuration. Based on the objective evaluation data (temperature, humidity, wind speed) collected above, and subjective evaluation data (questionnaire survey, user feedback) collected simultaneously, a comprehensive human comfort evaluation index and comfort function are constructed. Based on the obtained dynamic threshold ranges for energy consumption of existing electrical appliances in small buildings, the energy priority of electrical appliances in small buildings is dynamically adjusted. Combined with the comfort evaluation results, energy regulation strategies are selected and set. Based on the corresponding energy regulation strategies, the self-sufficiency and zero-carbon operation time of wind, solar, and energy storage small buildings is extended.

[0008] The present invention discloses a zero-carbon energy regulation method for wind-solar-storage buildings that takes into account the thermal comfort of personnel, comprising the following steps:

[0009] Step 1: Establish energy supply and consumption boundaries for small buildings equipped with independent microgrids, obtain the scale of small buildings, and construct the backend logical structure of a smart energy management system for small zero-carbon buildings, comprising four parts: data layer, computing layer, presentation layer, and application layer. Connect the data layer of the smart energy management system to a MySQL database to read energy supply data from wind, solar, and energy storage small buildings in real time at the energy supply end. Based on the determined energy consumption boundaries of occupants in small zero-carbon buildings, connect existing electrical appliances to smart sensors to collect energy consumption data of existing electrical appliances in the small buildings. Pre-enter energy supply technical parameters and material consumption data into the database, and obtain the energy and material consumption, savings, and substitution amounts through data processing. During the data collection and processing, use Lagrange interpolation to impute missing values.

[0010] Step 1.1: Based on the four aspects of personnel activities, machine operation, environmental damage and sudden energy consumption, make boundary determinations on the energy consumption behavior of small zero-carbon buildings in different scenarios;

[0011] Step 1.2: Construct the backend logical structure of a small-scale zero-carbon building smart energy management system, which includes four parts: data layer, computing layer, presentation layer, and application layer;

[0012] Step 1.3: Connect to the MySQL database in the data layer of the smart energy management system for small zero-carbon buildings, and read the energy supply data of wind, solar and energy storage small buildings in real time at the energy supply end;

[0013] Step 1.4: Based on the determined energy consumption boundaries of personnel behavior in small zero-carbon buildings, connect existing electrical appliances to smart sensors to collect subjective and objective evaluation data as well as energy consumption data of existing electrical appliances in the small building. This energy consumption data includes: ① the time nodes when existing indoor electrical appliances are turned on; ② the time nodes when existing indoor electrical appliances are turned off; ③ the usage duration of existing indoor electrical appliances; ④ the energy consumption of existing indoor electrical appliances. Connect the building's internal energy supply equipment to smart sensors to collect energy consumption data of non-existing electrical appliances in the small building, including: ⑤ the usage duration of electrical appliances carried by personnel and the amount of energy used by the building when these appliances are connected to the energy supply terminal of the small zero-carbon building. Install smart sensors on the building envelope to monitor energy losses caused by periodic environmental changes, including: ⑥ the time nodes when personnel enter the building; ⑦ the time nodes when personnel leave the building; ⑧ changes made by personnel to the building envelope (opening windows, opening doors); ⑩ energy losses of the building's energy storage system. Install smart sensors comprehensively at all energy-related locations inside the building to monitor activity data and sudden energy consumption.

[0014] Step 1.5: Enter energy supply technical parameters and material consumption data into the database in advance, and obtain the consumption, savings and substitution of energy and materials through data processing;

[0015] Step 1.6: During the data collection and processing, Lagrange interpolation is used to impute missing values ​​in order to handle possible anomalies and gaps in data, thereby reducing the impact on subsequent electricity consumption behavior analysis.

[0016] Step 2: Analyze the principal components of the activity data and determine the energy consumption behavior based on the PCA-DA comprehensive method. Classify the energy consumption behavior and determine the energy consumption of different categories of behavior to form a detailed activity dataset with clear feature distinctions, which will be used in Step 6 to accurately determine the energy consumption behavior in the building.

[0017] Step 2.1: Principal Component Analysis (PCA) is used to reduce the dimensionality of the dataset obtained in Step 1. The data is standardized, and the covariance matrix between features is calculated. Eigenvalue decomposition is performed on the covariance matrix to calculate eigenvalues ​​and eigenvectors. The eigenvalues ​​are sorted in descending order, and the eigenvectors corresponding to the k largest eigenvalues ​​are selected and used to form a dimensionality reduction matrix (k is a constant). The original data is multiplied by the dimensionality reduction matrix to output the dimensionality-reduced data, which includes four categories: the use of existing indoor electrical appliances (①-④), the use of electrical appliances carried by personnel (⑤⑥), energy loss due to physical activities of personnel (⑦-⑨), and energy loss from the building itself (⑩). The specific formula is as follows:

[0018] x ij =(x ij -mean(x j )) / std(x j )……………………………………(1)

[0019]

[0020] Cov(X)·V=V·D…………………………………………(3)

[0021] V k =[V1,V2,…,V k ]………………………………………………(4)

[0022] W = [V1,V2,…,V k ]………………………………………………(5)

[0023] Y=X·W………………………………………………………………(6)

[0024] Where, x ij The mean(x) represents the feature value of the j-th energy-using behavior of the i-th person. j std(x) represents the average value of the j-th energy-consuming behavior characteristic. j Let ) represent the standard deviation of the j-th energy-use behavior feature, n represent the number of samples, Cov(X) represent the covariance matrix, mean(X) represent the mean of the energy-use behavior dataset X, V represent the feature vector matrix, and D represent a diagonal matrix where the elements on the diagonal are the feature values ​​of energy-use behavior. k Y represents the selected feature vector matrix, X represents the dimensionality-reduced dataset, and W represents the dimensionality reduction matrix.

[0025] Step 2.2: Discriminant analysis (DA) is used to classify and analyze the activity data and determine energy consumption behavior. The fuzzy C-means algorithm (FCM) is used to classify the sample data D = {X1, X2, ..., X...}. n The clusters are divided into c clusters, namely C1, C2, ..., C6. c The dataset is grouped, and the cluster centers of each group are calculated to minimize the objective function value. By iteratively updating the cluster centers and membership matrix, the optimal cluster centers are obtained, thereby determining the energy consumption behavior of personnel. The specific formula is as follows:

[0026]

[0027]

[0028]

[0029]

[0030] Among them, J m Let c be the objective function for clustering energy use behavior, and u be the number of clusters formed by the selected energy use behavior dataset. ij x represents i Belonging to cluster center p j membership degree, u i The term is used to define the cluster centers, where m is a parameter used to control the degree of fuzziness in membership, and D is the number of clusters into which the data needs to be divided.

[0031] Step 3: Dynamically acquire the energy consumption dynamic thresholds of existing electrical appliances in small buildings that meet the natural endowments of wind and solar power based on deep learning. Based on the monitoring results of sudden energy consumption behavior, determine the occurrence cycle and energy consumption amount. Add an external parameter to the convolutional neural network (CNN) model to characterize the impact of sudden energy consumption behavior on energy management. Construct and train an LTMF-CNN deep learning model for small buildings, and use the trained LTMF-CNN deep learning model to predict the dynamic energy consumption thresholds of existing electrical appliances in small buildings.

[0032] Step 3.1: Preprocess the collected data on personnel energy consumption behavior and energy consumption, including data cleaning, outlier removal, and missing value imputation, to ensure data accuracy and completeness. Z-standardization is applied to the sample data using the following formula:

[0033] z-score=(X-mean(X)) / std(X)……………………(11)

[0034] Where z-score represents the deviation of each observation from the mean and standard deviation of the dataset, mean(X) represents the mean of dataset X, and std(X) represents the standard deviation of dataset X.

[0035] Step 3.2: Extract time features, appliance types, energy consumption of existing appliances, energy consumption of external appliances, personnel behavior, personnel habits, and hardware energy loss from the data, and establish a deep learning model to identify the changing patterns of energy consumption behavior;

[0036] Step 3.3: Divide the preprocessed energy consumption data into a training set and a test set, where the training set is used for model training and the test set is used for model evaluation;

[0037] Step 3.4: Construct a Convolutional Neural Network (CNN) deep learning model, including an input layer, convolutional layers, pooling layers, fully connected layers, and an output layer; the input layer incorporates the impact of sudden energy consumption behavior.

[0038] Step 3.5: Use the Mean Squared Error (MSE) loss function to measure the difference between the model's predictions and the actual values. The specific formula is as follows:

[0039]

[0040] Where L represents the loss value, N represents the number of samples, and Y represents the model output. Represents the actual value;

[0041] Step 3.6: Calculate the gradient using the backpropagation algorithm to update the weights and biases in the model, gradually bringing them closer to the optimal solution;

[0042] Step 3.7: Train and evaluate the LTMF-CNN deep learning model using training and test data. Input the collected data into the trained LTMF-CNN deep learning model and use the trained LTMF-CNN deep learning model to predict the dynamic energy consumption threshold of existing electrical appliances in small buildings.

[0043] Step 4: Based on the predicted dynamic energy consumption threshold of existing electrical appliances in small buildings, equip small buildings with energy storage batteries to cope with sudden energy supply fluctuations.

[0044] Step 4.1: Integrate future weather forecast data, when the light intensity is less than 1000W / m² 2 If the wind speed is less than 4 m / s, it is considered to be extreme weather. The battery reserves of small buildings should be able to meet the power demand under extreme weather conditions for 6 consecutive days.

[0045] Step 4.2: Based on the dynamic energy consumption threshold of electrical appliances obtained in Step 3.7, equip small wind-solar-storage buildings with energy storage batteries to cope with sudden energy supply fluctuations.

[0046] Step 5: Establish a comprehensive evaluation index for thermal comfort, PMV.

[0047] Step 5.1: Organize the datasets of subjective evaluations (questionnaires, user feedback) and objective evaluations (temperature, humidity, wind speed, etc.) to ensure that the evaluation indicators are measurable and comparable. Connect to air quality sensor data in real time to obtain indoor environmental parameters, including carbon dioxide concentration, temperature, and humidity.

[0048] Step 5.2: Construct the comprehensive evaluation index PMV, the specific formula is as follows:

[0049] PMV = (0.303e -0.036M +0.028){MW-3.05×10 -5 [5733-6.99(M-

[0050] W)-P a -0.42[MW-58.15]-1.7×10 -5 M(5867-P a -0.0014M(34-t) a -3.96×10 - 8 f cl h c (t cl -t a )}………………………………………………………………………………(13)

[0051]

[0052] t cl =35.7-0.028(MW)-I cl 3.96×10 -8 [(t cl +273) 4 -(t mrt +

[0053] 273) 4 ]+f cl h c (t cl -t a )}…………………………………………………(15)

[0054]

[0055] Where M represents the human body's energy metabolism rate, characterizing the amount of human activity; W represents the mechanical work performed by the human body, the mechanical efficiency of the human body's output under different activity intensities; P a t represents the partial pressure of water vapor in the air surrounding the human body. a t represents the air temperature of the environment surrounding the human body.mrt Indicates the average radiant temperature inside the room; I cl Indicates the thermal resistance of clothing; f cl t represents the surface area coefficient of clothing, which is the ratio of the actual surface area of ​​the human body when clothed to the surface area of ​​the naked body, and is related to the thermal resistance of the clothing; cl Indicates the outer surface temperature of the clothing; h c V represents the surface heat transfer coefficient; a Indicates air velocity.

[0056] Step 5.3: Define a comfort function. The goal of this comfort function is to minimize its value, that is, to get as close as possible to the set comfort temperature. The specific formula is as follows:

[0057]

[0058] Where T is the indoor air temperature, T set It is the set comfort temperature, ΔT cl It is the difference between human body thermal radiation and convective heat transfer, ΔT rad It is the difference between radiant temperature and indoor air temperature, ΔT air It is the difference between the indoor air temperature and the set temperature, ΔT met It represents the difference between heat generated by the human body and heat exchange with the environment, ΔT. ext It is the temperature difference between indoors and outdoors, c 1i c 2i c 3i c 4i c 5i is the weighting coefficient, used to adjust the relative impact of each parameter on comfort; N is the number of people involved in the calculation.

[0059] Step Six: Based on the dynamic threshold range of energy consumption of existing electrical appliances in small buildings, and combined with the comprehensive evaluation index PMV for thermal comfort, dynamically adjust the energy consumption priority of electrical appliances in small buildings. Based on the comfort evaluation results, select and set energy regulation strategies, and extend the self-sufficiency and zero-carbon operation time of wind, solar and energy storage small buildings based on the corresponding energy regulation strategies.

[0060] Step 6.1: The objective is to minimize energy consumption and the comfort function. The specific formula is as follows:

[0061] Obj=w1•(Energy-Energy_desired) 2 +w2·f………………(18)

[0062] Where Energy represents actual energy consumption, Energy_desired represents desired energy consumption; f represents the comfort function, whose value approaches the minimum, that is, to the greatest extent possible to get closer to the set comfort level; w1 and w2 are weighting coefficients used to balance the importance of different objectives.

[0063] Step 6.2: Considering the energy demand, energy consumption limitations, and environmental conditions of electrical appliances, based on the dynamic threshold range of energy consumption of existing electrical appliances in small buildings, dynamically adjust the energy priority of electrical appliances in small buildings. Combined with the comfort evaluation results, select and set energy regulation strategies. Based on the corresponding energy regulation strategies, extend the self-sufficiency and zero-carbon operation time of wind, solar, and energy storage small buildings. The specific implementation method is as follows:

[0064] The energy demand of electrical appliances shall not exceed the set energy supply capacity:

[0065] ∑ i E i ≤E_supply…………………………………………(19) The relationship between the energy demand of electrical appliances and their priority weights:

[0066] E i ≤w i ·E max ……………………………………………(20)

[0067] Among them, E i E_supply represents the energy demand of appliance i, E_supply represents the energy supply capacity of the small building, and w i E represents the priority weight of small building electrical appliance i. max This indicates the maximum energy demand of small building electrical appliances.

[0068] The energy consumption of electrical appliances should not exceed their energy requirements:

[0069] C i ≤E i ……………………………………………………(twenty one)

[0070] Among them, C i This represents the energy consumption of appliance i.

[0071] Under specific environmental conditions, the operating parameters of electrical appliances need to be controlled within a certain range:

[0072] P min ≤P i ≤P max …………………………………………(twenty two)

[0073] T min≤T i ≤T max …………………………………………(twenty three)

[0074] Among them, P i T represents the power of appliance i. i This indicates the temperature of appliance i.

[0075] The interrelationship between the energy consumption of electrical appliances:

[0076] C i ≥∑ j C ij ……………………………………………………(twenty four)

[0077] Among them, C ij This represents the coordinated energy consumption between appliance i and appliance j.

[0078] The constraints take into account the dynamic changes in the energy storage state:

[0079] E(t+1)=E(t)+C_charge………………………………(25)

[0080] Electricity demand must be met before an emergency occurs:

[0081] ∑E_demand(t)≤∑E_supply(t)t=1,2,..,1s-1……………………(26) Balancing electricity demand during emergencies:

[0082] ∑E_demand(t)≤∑E_supply(t),t=s……………………(27) Power supply meets demand in the event of an emergency:

[0083] ∑E_demand(t)≤∑E_supply(t)t=s+1,s+2,……………………(28)

[0084] Where t is the time step, the time of the emergency is t = s, C_charge represents the rate of change of energy storage, E_supply represents the amount of electricity supplied, and E_demand represents the amount of electricity demanded.

[0085] The small building described in this invention specifically refers to a small building that does not have a thermal energy system.

[0086] Beneficial effects:

[0087] 1. The energy regulation method for zero-carbon operation of wind, solar and energy storage buildings that considers human thermal comfort disclosed in this invention dynamically adjusts the energy consumption priority of electrical appliances in small buildings based on the dynamic threshold range of energy consumption of existing electrical appliances in small buildings while ensuring human comfort. Combined with the comfort evaluation results, an energy regulation strategy is selected and set, and the duration of autonomous energy supply and zero-carbon operation of wind, solar and energy storage small buildings is extended based on the corresponding energy regulation strategy.

[0088] 2. The zero-carbon operation energy regulation method for wind, solar and energy storage buildings that considers the thermal comfort of personnel disclosed in this invention constructs a personnel comfort evaluation index and a comfort function based on subjective and objective evaluation data, quantifies personnel comfort, and ensures that the level of personnel comfort does not decrease during the energy regulation process aimed at minimizing energy consumption.

[0089] 3. The zero-carbon operation energy regulation method for wind, solar and energy storage buildings that considers the thermal comfort of people disclosed in this invention is based on intelligent sensors and deep learning algorithms. It performs dimensionality reduction, clustering and prediction on the collected dynamic energy consumption data. Using the constructed LTMF-CNN deep learning model, it effectively obtains the energy consumption change pattern of electrical appliances in the building and predicts the future dynamic threshold of energy consumption. At the same time, combined with the accessed meteorological data, it effectively reduces the amount of energy storage equipment used, improves the health and service life of energy storage equipment, and effectively ensures that even if small buildings of wind, solar and energy storage microgrids cannot generate electricity themselves under extreme weather conditions, they can still provide continuous energy supply. Attached Figure Description

[0090] Figure 1 A schematic diagram of a zero-carbon energy regulation method for small wind-solar-storage buildings that takes into account the thermal comfort of personnel, according to the present invention. Detailed Implementation

[0091] To better understand the above-mentioned objectives, features, and advantages of the present invention, the following detailed description is provided using the embodiment of a "zero-carbon cabin" as an example. This cabin has a floor area of ​​36 square meters, a length and width of 6 meters, and a height of 2.5 meters. It utilizes a standardized high-strength cold-formed thin-walled steel system, allowing for disassembly, relocation, and recycling. The cabin utilizes two 500W fans on the roof and eight 450W monocrystalline silicon wafers to supply clean energy, meeting the electricity needs of indoor air conditioning, office equipment, and the external wall screen. It is also equipped with a heat recovery system and an energy storage system, achieving energy self-sufficiency and intelligent control, thus achieving zero carbon emissions.

[0092] like Figure 1 As shown in the figure, this embodiment discloses a zero-carbon operation energy regulation method for wind-solar-storage buildings that considers the thermal comfort of personnel. The specific implementation steps are as follows:

[0093] Step 1: Establish energy supply and consumption boundaries for small buildings equipped with independent microgrids, obtain the scale of small buildings, and construct the backend logical structure of a smart energy management system for small zero-carbon buildings, comprising four parts: data layer, computing layer, presentation layer, and application layer. Connect the data layer of the smart energy management system to a MySQL database to read energy supply data from wind, solar, and energy storage small buildings in real time at the energy supply end. Based on the determined energy consumption boundaries of occupants in small zero-carbon buildings, connect existing electrical appliances to smart sensors to collect energy consumption data of existing electrical appliances in the small buildings. Pre-enter energy supply technical parameters and material consumption data into the database, and obtain the energy and material consumption, savings, and substitution amounts through data processing. During the data collection and processing, use Lagrange interpolation to impute missing values.

[0094] Step 1.1: Based on the four aspects of personnel activities, machine operation, environmental damage and sudden energy consumption, make boundary determinations on the energy consumption behavior of small zero-carbon buildings in different scenarios;

[0095] Step 1.2: Construct the backend logical structure of a small-scale zero-carbon building smart energy management system, which includes four parts: data layer, computing layer, presentation layer, and application layer;

[0096] Step 1.3: Connect to the MySQL database in the data layer of the smart energy management system for small zero-carbon buildings, and read the energy supply data of wind, solar and energy storage small buildings in real time at the energy supply end;

[0097] Step 1.4: Based on the energy consumption boundary of personnel behavior in the small zero-carbon building as determined, connect the existing electrical appliances to the smart sensor to collect subjective and objective evaluation data as well as energy consumption data of the existing electrical appliances in the small building;

[0098] Step 1.5: Enter energy supply technical parameters and material consumption data into the database in advance, and obtain the consumption, savings and substitution of energy and materials through data processing;

[0099] Step 1.6: During the data collection and processing, Lagrange interpolation is used to impute missing values ​​in order to handle possible anomalies and gaps in data, thereby reducing the impact on subsequent electricity consumption behavior analysis.

[0100] According to an embodiment of the present invention, step 1.1 specifically involves: making boundary determinations on the energy consumption behavior of small zero-carbon buildings in scenarios such as sunny days, light intensity less than 1000W / m^2, and wind speed less than 4m / s.

[0101] According to an embodiment of the present invention, step 1.4 specifically includes: the connected electrical appliances include: lights, displays, office computers, fresh air systems, and air conditioners; the collected energy consumption data of the electrical appliances includes ① the opening time of the existing indoor electrical appliances, ② the closing time of the existing indoor electrical appliances, ③ the usage time of the existing indoor electrical appliances, and ④ the energy consumption of the existing indoor electrical appliances; connecting the building's internal energy supply equipment to smart sensors to collect energy consumption data of non-existent electrical appliances in the small building, including ⑤ the usage time of electrical appliances carried by personnel and the amount of building energy used when the electrical appliances carried by personnel are connected to the energy supply terminal of the small zero-carbon building; installing smart sensors on the building envelope to monitor energy loss caused by periodic environmental changes, including ⑥ the time of personnel entering the building, ⑦ the time of personnel leaving the building, ⑧ changes made by personnel to the building envelope (opening windows, opening doors), and ⑩ energy loss of the building energy storage system; and installing smart sensors comprehensively in all locations involving energy consumption inside the building to monitor activity data and sudden energy consumption.

[0102] According to an embodiment of the present invention, the technical data input in step 1.5 includes: ① Lighting: power 0.09kW, duration 6h, energy consumption 0.53kWh; ② Display screen: power 0.25kW, duration 10h, energy consumption 2.5kWh; ③ Office computer: power 0.4kW, duration 4h, energy consumption 1.6kWh; ④ Fresh air system: power 0.09kW, duration 10h, energy consumption 0.85kWh; ⑤ Air conditioner: power 0.36kW, duration 10h, energy consumption 3.6kWh.

[0103] Step 2: Analyze the principal components of the activity data and determine the energy consumption behavior based on the PCA-DA comprehensive method. Classify the energy consumption behavior and determine the energy consumption of different categories of behavior to form a detailed activity dataset with clear feature distinctions, which will be used in Step 6 to accurately determine the energy consumption behavior in the building.

[0104] Step 2.1: Principal Component Analysis (PCA) is used to reduce the dimensionality of the dataset obtained in Step 1;

[0105] Step 2.2: Use discriminant analysis (DA) to classify and analyze the activity data and determine energy consumption behavior.

[0106] According to an embodiment of the present invention, step 2.1 specifically involves: standardizing the data, calculating the covariance matrix between features, performing eigenvalue decomposition on the covariance matrix, calculating eigenvalues ​​and eigenvectors, sorting the eigenvalues ​​in descending order, selecting and retaining the eigenvectors corresponding to the k largest eigenvalues ​​to form a dimensionality reduction matrix (k is a constant), multiplying the original data with the dimensionality reduction matrix, and outputting the dimensionality-reduced data. The dimensionality-reduced data includes four categories: the use of existing indoor electrical appliances (①-④), the use of electrical appliances carried by personnel (⑤⑥), energy loss caused by personnel's physical activities (⑦-⑨), and energy loss of the building itself (⑩). The specific calculation formula is as follows:

[0107] x ij =(x ij -mean(x j )) / std(x j )

[0108]

[0109] Cov(X)·V=V·D

[0110] V k =[V1,V2,…,V k ]

[0111] W = [V1,V2,…,V k ]

[0112] Y = X·W

[0113] Where, x ij The mean(x) represents the feature value of the j-th energy-using behavior of the i-th person. j std(x) represents the average value of the j-th energy-consuming behavior characteristic. j Let ) represent the standard deviation of the j-th energy-use behavior feature, n represent the number of samples, Cov(X) represent the covariance matrix, mean(X) represent the mean of the energy-use behavior dataset X, V represent the feature vector matrix, and D represent a diagonal matrix where the elements on the diagonal are the feature values ​​of energy-use behavior. k Y represents the selected feature vector matrix, X represents the dimensionality-reduced dataset, and W represents the dimensionality reduction matrix.

[0114] According to an embodiment of the present invention, step 2.2 specifically involves: using the fuzzy C-means algorithm (FCM) to process the sample data D = {X1, X2, ..., X...} n The clusters are divided into c clusters, namely C1, C2, ..., Cc. cThe dataset is grouped, and the cluster centers of each group are calculated to minimize the objective function value. By iteratively updating the cluster centers and membership matrix, the optimal cluster centers are obtained, thereby determining the energy consumption behavior of personnel. The specific formula is as follows:

[0115]

[0116]

[0117]

[0118]

[0119] Among them, J m Let c be the objective function for clustering energy use behavior, and u be the number of clusters formed by the selected energy use behavior dataset. ij x represents i Belonging to cluster center p j membership degree, u i The term is used to define the cluster centers, where m is a parameter used to control the degree of fuzziness in membership, and D is the number of clusters into which the data needs to be divided.

[0120] Step 3: Dynamically acquire the energy consumption dynamic thresholds of existing electrical appliances in small buildings that meet the natural endowments of wind and solar power based on deep learning. Based on the monitoring results of sudden energy consumption behavior, determine the occurrence cycle and energy consumption amount. Add an external parameter to the convolutional neural network (CNN) model to characterize the impact of sudden energy consumption behavior on energy management. Construct and train an LTMF-CNN deep learning model for small buildings, and use the trained LTMF-CNN deep learning model to predict the dynamic energy consumption thresholds of existing electrical appliances in small buildings.

[0121] Step 3.1: Preprocess the collected data on personnel energy consumption behavior and energy consumption, including data cleaning, handling outliers, and filling in missing values, to ensure the accuracy and completeness of the data;

[0122] Step 3.2: Extract time features, appliance types, energy consumption of existing appliances, energy consumption of external appliances, personnel behavior, personnel habits, and hardware energy loss from the data, and establish a deep learning model to identify the changing patterns of energy consumption behavior;

[0123] Step 3.3: Divide the preprocessed energy consumption data into a training set and a test set, where the training set is used for model training and the test set is used for model evaluation;

[0124] Step 3.4: Construct a Convolutional Neural Network (CNN) deep learning model, including an input layer, convolutional layers, pooling layers, fully connected layers, and an output layer; the input layer incorporates the impact of sudden energy consumption behavior.

[0125] Step 3.5: Use the mean squared error (MSE) loss function to measure the difference between the model's predictions and the actual values;

[0126] Step 3.6: Calculate the gradient using the backpropagation algorithm to update the weights and biases in the model, gradually bringing them closer to the optimal solution;

[0127] Step 3.7: Train and evaluate the LTMF-CNN deep learning model using training and test data. Input the collected data into the trained LTMF-CNN deep learning model and use the trained LTMF-CNN deep learning model to predict the dynamic energy consumption threshold of existing electrical appliances in small buildings.

[0128] According to an embodiment of the present invention, step 3.1 specifically involves: performing Z-standardization on the sample data, using the following formula:

[0129] z-score = (X-mean(X)) / std(X)

[0130] Where z-score represents the deviation of each observation from the mean and standard deviation of the dataset, mean(X) represents the mean of dataset X, and std(X) represents the standard deviation of dataset X.

[0131] According to an embodiment of the present invention, the specific formula for the mean square error in step 3.5 is as follows:

[0132]

[0133] Where L represents the loss value, N represents the number of samples, and Y represents the model output. Represents the actual value.

[0134] Step 4: Based on the predicted dynamic energy consumption threshold of existing electrical appliances in small buildings, equip small buildings with energy storage batteries to cope with sudden energy supply fluctuations.

[0135] Step 4.1: Access future weather forecast data. The battery reserves of small buildings should be able to meet the power demand under extreme weather conditions for 6 consecutive days.

[0136] Step 4.2: Based on the dynamic energy consumption threshold of electrical appliances obtained in Step 3.7, equip small wind-solar-storage buildings with energy storage batteries to cope with sudden energy supply fluctuations.

[0137] According to an embodiment of the present invention, step 4.1 specifically involves: accessing future weather forecast data, when the light intensity is less than 1000 W / m². 2 If the wind speed is less than 4 m / s, it is considered to be extreme weather.

[0138] According to an embodiment of the present invention, step 4.2 specifically involves equipping the small wind-solar-storage building with 24 100AH ​​lithium batteries, which can store 28 kWh of electricity.

[0139] Step 5: Establish a comprehensive evaluation index for thermal comfort, PMV.

[0140] Step 5.1: Organize the datasets of subjective evaluations (questionnaires, user feedback) and objective evaluations (temperature, humidity, wind speed, etc.) to ensure that the evaluation indicators are measurable and comparable. Connect to air quality sensor data in real time to obtain indoor environmental parameters, including carbon dioxide concentration, temperature, and humidity.

[0141] Step 5.2: Construct a comprehensive evaluation index, PMV;

[0142] Step 5.3: Define a comfort function. The goal of this comfort function is to minimize its value, that is, to get as close as possible to the set comfort temperature.

[0143] According to an embodiment of the present invention, step 5.1 specifically involves: constructing a comprehensive evaluation index (PMV), the specific formula of which is:

[0144] PMV = (0.303e -0.036M +0.028){MW-3.05×10 -5 [5733-6.99(MW)-P a -0.42[MW-58.15]-1.7×10 -5 M(5867-P a -0.0014M(34-t) a -3.96×10 -8 f cl h c (t cl -t a )}

[0145]

[0146] t cl =35.7-0.028(MW)-I cl 3.96×10 -8 [(t cl +273) 4 -(t mrt +273) 4 ]+f cl h c (t cl -t a )}

[0147]

[0148] Where M represents the human body's energy metabolism rate, characterizing the amount of human activity; W represents the mechanical work performed by the human body, the mechanical efficiency of the human body's output under different activity intensities; P a t represents the partial pressure of water vapor in the air surrounding the human body. a t represents the air temperature of the environment surrounding the human body. mrt Indicates the average radiant temperature inside the room; I cl Indicates the thermal resistance of clothing; f cl t represents the surface area coefficient of clothing, which is the ratio of the actual surface area of ​​the human body when clothed to the surface area of ​​the naked body, and is related to the thermal resistance of the clothing; cl Indicates the outer surface temperature of the clothing; h c V represents the surface heat transfer coefficient; a Indicates air velocity.

[0149] According to an embodiment of the present invention, step 5.3 specifically involves defining a comfort function, the specific formula of which is:

[0150]

[0151] Where T is the indoor air temperature, T set It is the set comfort temperature, ΔT cl It is the difference between human body thermal radiation and convective heat transfer, ΔT rad It is the difference between radiant temperature and indoor air temperature, ΔT air It is the difference between the indoor air temperature and the set temperature, ΔT met It represents the difference between heat generated by the human body and heat exchange with the environment, ΔT. ext It is the temperature difference between indoors and outdoors, c 1i c 2i c 3i c 4i c 5i is the weighting coefficient, used to adjust the relative impact of each parameter on comfort, and N is the number of people involved in the calculation.

[0152] Step Six: Based on the dynamic threshold range of energy consumption of existing electrical appliances in small buildings, and combined with the comprehensive evaluation index PMV for thermal comfort, dynamically adjust the energy consumption priority of electrical appliances in small buildings. Based on the comfort evaluation results, select and set energy regulation strategies, and extend the self-sufficiency and zero-carbon operation time of wind, solar and energy storage small buildings based on the corresponding energy regulation strategies.

[0153] Step 6.1: The objectives are to minimize energy consumption and minimize the comfort function;

[0154] Step 6.2: Considering the energy demand, energy consumption limits and environmental conditions of electrical appliances, dynamically adjust the energy priority of electrical appliances in small buildings based on the dynamic threshold range of energy consumption of existing electrical appliances in small buildings. Combined with the comfort evaluation results, select and set energy regulation strategies, and extend the self-sufficiency and zero-carbon operation time of wind, solar and storage small buildings based on the corresponding energy regulation strategies.

[0155] According to an embodiment of the present invention, step 6.1 specifically involves minimizing energy consumption and minimizing the comfort function as objectives, with the specific formula being:

[0156] Obj=w1·(Energy-Energy_desired) 2 +w2·f

[0157] Where Energy represents actual energy consumption, Energy_desired represents desired energy consumption; f represents the comfort function, whose value approaches the minimum, that is, to the greatest extent possible to get closer to the set comfort level; w1 and w2 are weighting coefficients used to balance the importance of different objectives.

[0158] According to an embodiment of the present invention, step 6.2 specifically involves: considering the energy demand, energy consumption limitations, and environmental conditions of electrical appliances, dynamically adjusting the energy consumption priority of electrical appliances in small buildings based on the existing dynamic threshold range of energy consumption of electrical appliances in small buildings, selecting and setting energy regulation strategies based on comfort evaluation results, and extending the self-sufficiency and zero-carbon operation time of wind, solar, and energy storage small buildings based on the corresponding energy regulation strategies. The specific constraint formula is as follows:

[0159] The energy demand of electrical appliances shall not exceed the set energy supply capacity:

[0160]

[0161] The relationship between the energy demand of electrical appliances and their priority weights:

[0162] E i ≤w i ·E max

[0163] Among them, E i E_supply represents the energy demand of appliance i, E_supply represents the energy supply capacity of the small building, and w i E represents the priority weight of small building electrical appliance i. max This indicates the maximum energy demand of small building electrical appliances.

[0164] The energy consumption of electrical appliances should not exceed their energy requirements:

[0165] C i ≤E i

[0166] Among them, C i This represents the energy consumption of appliance i.

[0167] Under specific environmental conditions, the operating parameters of electrical appliances need to be controlled within a certain range:

[0168] P min ≤P i ≤P max

[0169] T min ≤T i ≤T max

[0170] Among them, P i T represents the power of appliance i. i This indicates the temperature of appliance i.

[0171] The interrelationship between the energy consumption of electrical appliances:

[0172]

[0173] Among them, C ij This represents the coordinated energy consumption between appliance i and appliance j.

[0174] The constraints take into account the dynamic changes in the energy storage state:

[0175] E(t+1)=E(t)+C_charge

[0176] Electricity demand must be met before an emergency occurs:

[0177] ∑E_demand(t)≤∑E_supply(t)t=1,2,..,1s-1

[0178] Balancing electricity demand during emergencies:

[0179] ∑E_demand(t)≤∑E_supply(t),t=s

[0180] Power supply meets demand in the event of an emergency:

[0181] ∑E_demand(t)≤∑E_supply(t)t=s+1,s+2,…

[0182] Where t is the time step, the time of the emergency is t = s, C_charge represents the rate of change of energy storage, E_supply represents the amount of electricity supplied, and E_demand represents the amount of electricity demanded.

[0183] Example results:

[0184] 1. In this embodiment, while ensuring human comfort, the energy consumption priority of electrical appliances in small buildings is dynamically adjusted based on the dynamic threshold range of energy consumption of existing electrical appliances in small buildings. Combined with the comfort evaluation results, an energy regulation strategy is selected and set. Based on the corresponding energy regulation strategy, the self-powered and zero-carbon operation time of wind, solar and energy storage small buildings is extended.

[0185] 2. Based on subjective and objective evaluation data, this embodiment constructs personnel comfort evaluation indicators and comfort functions to quantify personnel comfort and ensure that the level of personnel comfort does not decrease during the energy regulation process aimed at minimizing energy consumption.

[0186] 3. This embodiment uses intelligent sensors and deep learning algorithms to perform dimensionality reduction, clustering, and prediction on the collected dynamic energy consumption data. By using the constructed LTMF-CNN deep learning model, it effectively obtains the energy consumption change patterns of electrical appliances in buildings and predicts future dynamic energy consumption thresholds. At the same time, combined with the accessed meteorological data, it effectively reduces the amount of energy storage equipment used, improves the health and service life of energy storage equipment, and effectively ensures that even if small buildings in the wind, solar, and energy storage microgrid cannot generate their own electricity under extreme weather conditions, they can still continue to supply energy.

[0187] 4. The "zero-carbon cabin" in the embodiment achieves self-sufficiency for 50 days without external power supply, and overcomes poor wind and light conditions for up to 6 days. This is better than the effect of the same weather and hardware conditions in this embodiment, where the invention can only operate continuously for a maximum of 2 days without the invention.

Claims

1. A zero-carbon energy regulation method for wind-solar-storage buildings considering human thermal comfort, characterized by: Includes the following steps, Step 1: Establish energy supply and consumption boundaries for small buildings equipped with independent microgrids, obtain the scale of small buildings, and construct the backend logical structure of a smart energy management system for small zero-carbon buildings, comprising four parts: data layer, computing layer, display layer, and application layer. Connect a MySQL database to the data layer of the smart energy management system for small zero-carbon buildings, and read the energy supply data of wind, solar, and energy storage small buildings in real time at the energy supply end. Based on the determined energy consumption boundaries of personnel behavior in small zero-carbon buildings, connect existing electrical appliances to smart sensors to collect energy consumption data of existing electrical appliances in small buildings. Pre-enter energy supply technical parameters and material consumption data into the database, and obtain the energy and material consumption, savings, and substitution amounts through data processing. During the data collection and processing process, use Lagrange interpolation to impute missing values. Step 2: Analyze the principal components of the activity data and determine the energy consumption behavior based on the PCA-DA synthesis method. Classify the energy consumption behavior and determine the energy consumption of different categories of behavior to form an activity dataset with clear feature distinctions. Step 3: Based on deep learning, dynamically obtain the energy consumption dynamic threshold of existing electrical appliances in small buildings that meet the natural endowment of wind and light; based on the monitoring results of sudden energy consumption behavior, determine the occurrence cycle and energy consumption, add an external parameter to the convolutional neural network (CNN) model to characterize the impact of sudden energy consumption behavior on energy management, construct and train an LTMF-CNN deep learning model for small buildings, and use the trained LTMF-CNN deep learning model to predict the energy consumption dynamic threshold of existing electrical appliances in small buildings. Step 4: Based on the predicted dynamic energy consumption threshold of existing electrical appliances in small buildings, equip small buildings with energy storage batteries to cope with sudden energy supply fluctuations. Step 5: Establish a comprehensive evaluation index for thermal comfort, PMV; Step five is implemented as follows: Step 5.1: Utilize subjective and objective evaluation datasets to ensure that the evaluation indicators are measurable and comparable. Connect to air quality sensor data in real time to obtain indoor environmental parameters, including carbon dioxide concentration, temperature, and humidity. Step 5.2: Construct the comprehensive evaluation index PMV, the specific formula is as follows: ………………………………………………………………………………(13) ………………………………………………………………………… (15) in, It represents the human body's energy metabolism rate and indicates the amount of human activity. It represents the mechanical work done by the human body, and the mechanical efficiency output by the human body under different activity intensities; This indicates the partial pressure of water vapor in the air surrounding the human body. Indicates the air temperature of the environment surrounding the human body; This indicates the average radiant temperature inside the room. Indicates the thermal resistance of the clothing; The surface area coefficient of clothing is the ratio of the actual surface area of ​​the human body after being clothed to the surface area of ​​the naked human body, and it is related to the thermal resistance of clothing. Indicates the temperature of the outer surface of the clothing; Indicates the surface heat transfer coefficient; Indicates air velocity; Step 5.3: Define a comfort function. The goal of this comfort function is to minimize its value, that is, to get as close as possible to the set comfort temperature. The specific formula is as follows: …………………………………………………………………………… (17) in, It is the indoor air temperature. It is the set comfort temperature. It is the difference between human body thermal radiation and convective heat transfer. It is the difference between radiant temperature and indoor air temperature. It is the difference between the indoor air temperature and the set temperature. It represents the difference between metabolic heat and ambient heat exchange. It is the temperature difference between indoors and outdoors. , , , , is the weighting coefficient, used to adjust the relative impact of each parameter on comfort; N is the number of people participating in the calculation. Step Six: Based on the dynamic threshold range of energy consumption of existing electrical appliances in small buildings, and combined with the comprehensive evaluation index PMV for thermal comfort, dynamically adjust the energy consumption priority of electrical appliances in small buildings. Based on the comfort evaluation results, select and set energy regulation strategies, and extend the self-sufficiency and zero-carbon operation time of wind, solar and energy storage small buildings based on the corresponding energy regulation strategies.

2. The energy regulation method for zero-carbon operation of wind-solar-storage buildings considering human thermal comfort as described in claim 1, characterized in that: The implementation method for step one is as follows: Step 1.1: Based on the four aspects of personnel activities, machine operation, environmental damage and sudden energy consumption, make boundary determinations on the energy consumption behavior of small zero-carbon buildings in different scenarios; Step 1.2: Construct the backend logical structure of a small-scale zero-carbon building smart energy management system, which includes four parts: data layer, computing layer, presentation layer, and application layer; Step 1.3: Connect to the MySQL database in the data layer of the smart energy management system for small zero-carbon buildings, and read the energy supply data of wind, solar and energy storage small buildings in real time at the energy supply end; Step 1.4: Based on the energy consumption boundary of personnel behavior in the small zero-carbon building as determined, connect the existing electrical appliances to the smart sensor to collect subjective and objective evaluation data as well as the energy consumption data of the existing electrical appliances in the small building. The energy consumption data of the appliances includes ① the time node when the existing indoor appliances are turned on, ② the time node when the existing indoor appliances are turned off, ③ the usage time of the existing indoor appliances, and ④ the energy consumption of the existing indoor appliances. Connect the building's internal energy supply equipment to smart sensors to collect energy consumption data of non-existent electrical appliances in small buildings, including ⑤ the usage time of electrical appliances carried by people and the amount of building energy used when electrical appliances carried by people are connected to the energy supply terminal of small zero-carbon buildings; install smart sensors on the building envelope to monitor energy loss caused by periodic environmental changes, including ⑥ the time when people enter the building, ⑦ the time when people leave the building, ⑧ changes made by people to the building envelope, and ⑩ energy loss of the building energy storage system. Smart sensors are installed throughout the building at all locations where energy is used to monitor activity data and sudden energy consumption. Step 1.5: Enter energy supply technical parameters and material consumption data into the database in advance, and obtain the consumption, savings and substitution of energy and materials through data processing; Step 1.6: During the data collection and processing, Lagrange interpolation is used to impute missing values ​​in order to handle possible outliers and missing data.

3. The energy regulation method for zero-carbon operation of wind-solar-storage buildings considering human thermal comfort as described in claim 2, characterized in that: The second step is implemented as follows: Step 2.1: Principal Component Analysis (PCA) is used to reduce the dimensionality of the dataset obtained in Step 1. The data is standardized, and the covariance matrix between features is calculated. Eigenvalue decomposition is performed on the covariance matrix, calculating eigenvalues ​​and eigenvectors. The eigenvalues ​​are sorted in descending order, and the eigenvectors corresponding to the k largest eigenvalues ​​are selected and used to form the dimensionality reduction matrix, where k is a constant. The original data is multiplied by the dimensionality reduction matrix to output the dimensionality-reduced data, which includes four categories: the use of existing indoor electrical appliances ①-④, the use of electrical appliances carried by personnel ⑤⑥, energy loss due to physical activities of personnel ⑦-⑨, and energy loss from the building itself ⑩. The specific formula is as follows: ……………………………………(1) ………………(2) …………………………………………………(3) ………………………………………………(4) ………………………………………………(5) …………………………………………………………(6) in, Indicates the first The first person The characteristic value of energy consumption behavior Indicates the first The average value of each energy-consuming behavior characteristic. Indicates the first Standard deviation of energy-consuming behavior characteristics Indicates the number of samples. Represents the covariance matrix. Data set representing human energy consumption behavior The mean, Represents the eigenvector matrix, Let represent a diagonal matrix, where the elements on the diagonal are the eigenvalues ​​of the energy behavior. This represents the selected eigenvector matrix. This represents the dataset after dimensionality reduction. Represents the original dataset. Represents a dimension-reduced matrix; Step 2.2: Discriminant analysis (DA) is used to classify and analyze the activity data and determine energy-consuming behaviors. The fuzzy C-means algorithm (FCM) is then used to analyze the sample data. Divided into Clusters, namely The dataset is grouped, and the cluster centers of each group are calculated to minimize the objective function value. By iteratively updating the cluster centers and membership matrix, the optimal cluster centers are obtained, thereby determining the energy consumption behavior of personnel. The specific formula is as follows: …………………………(7) ………………………………(8) ……………………………………(9) ……………………………………(10) in, The objective function for clustering personnel energy use behavior is... This indicates the number of clusters formed by the selected energy behavior dataset. express Belonging to the cluster center membership degree As the class center, This refers to the parameter used to control the degree of fuzziness in membership. This refers to the number of clusters into which the data needs to be divided.

4. The energy regulation method for zero-carbon operation of wind-solar-storage buildings considering human thermal comfort as described in claim 3, characterized in that: The method for implementing step three is as follows: Step 3.1: Preprocess the collected data on personnel energy consumption behavior and energy consumption, including data cleaning, outlier removal, and missing value imputation to ensure data accuracy and completeness; perform Z-standardization on the sample data, using the following formula: …………………………(11) in, This represents the deviation of each observation from the mean and standard deviation of the dataset. Represents the dataset The average value, This represents the standard deviation of dataset X; Step 3.2: Extract time features, appliance types, energy consumption of existing appliances, energy consumption of external appliances, personnel behavior, personnel habits, and hardware energy loss from the data, and establish a deep learning model to identify the changing patterns of energy consumption behavior; Step 3.3: Divide the preprocessed energy consumption data into a training set and a test set, where the training set is used for model training and the test set is used for model evaluation; Step 3.4: Construct a Convolutional Neural Network (CNN) deep learning model, including an input layer, convolutional layers, pooling layers, fully connected layers, and an output layer; the input layer incorporates the impact of sudden energy consumption behavior. Step 3.5: Use the Mean Squared Error (MSE) loss function to measure the difference between the model's predictions and the actual values. The specific formula is as follows: …………………………………………(12) in, Indicates the loss value. Indicates the number of samples. This represents the output of the model. Represents the actual value; Step 3.6: Calculate the gradient using the backpropagation algorithm to update the weights and biases in the model, gradually bringing them closer to the optimal solution; Step 3.7: Train and evaluate the LTMF-CNN deep learning model using training and test data. Input the collected data into the trained LTMF-CNN deep learning model and use the trained LTMF-CNN deep learning model to predict the dynamic energy consumption threshold of existing electrical appliances in small buildings.

5. The energy regulation method for zero-carbon operation of wind-solar-storage buildings considering human thermal comfort as described in claim 4, characterized in that: Step four is implemented as follows: Step 4.1: Integrate future weather forecast data, when the light intensity is less than 1000. Wind speed less than 4 It is believed that in the event of extreme weather, the battery reserves of small buildings should be sufficient to meet the electricity demand for six consecutive days under extreme weather conditions. Step 4.2: Based on the dynamic energy consumption threshold of electrical appliances obtained in Step 3.7, equip small wind-solar-storage buildings with energy storage batteries to cope with sudden energy supply fluctuations.

6. The energy regulation method for zero-carbon operation of wind-solar-storage buildings considering human thermal comfort as described in claim 5, characterized in that: Step six is ​​implemented as follows: Step 6.1: The objective is to minimize energy consumption and the comfort function. The specific formula is as follows: ………………(18) in, Indicates actual energy consumption. For expected energy consumption; This represents a comfort function whose value approaches its minimum, meaning it gets as close as possible to the set comfort level. and These are weighting coefficients used to balance the importance of different objectives; Step 6.2: Considering the energy demand, energy consumption limitations, and environmental conditions of electrical appliances, based on the dynamic threshold range of energy consumption of existing electrical appliances in small buildings, dynamically adjust the energy priority of electrical appliances in small buildings. Combined with the comfort evaluation results, select and set energy regulation strategies. Based on the corresponding energy regulation strategies, extend the self-sufficiency and zero-carbon operation time of wind, solar, and energy storage small buildings. The specific implementation method is as follows: The energy demand of electrical appliances shall not exceed the set energy supply capacity: …………………………………………(19) The relationship between the energy demand of electrical appliances and their priority weights: ……………………………………………(20) in, Indicates electrical appliances energy demand, Indicates the energy supply capacity of small buildings. Indicates small building electrical appliances Priority weights, This indicates the maximum energy demand for small building electrical appliances; The energy consumption of electrical appliances should not exceed their energy requirements: ……………………………………………………(21) in, Indicates electrical appliances Energy consumption; Under specific environmental conditions, the operating parameters of electrical appliances need to be controlled within a certain range: …………………………………………(22) …………………………………………(23) in, Indicates electrical appliances power, Indicates electrical appliances Temperature; The interrelationship between the energy consumption of electrical appliances: ……………………………………………………(24) in, Indicates electrical appliances With electrical appliances Coordinated energy consumption between The constraints take into account the dynamic changes in the energy storage state: ……………………………………(25) Electricity demand must be met before an emergency occurs: t=1,2,..,1s-1……………………(26) Balancing electricity demand during emergencies: , t=s………………………………(27) Power supply meets demand in the event of an emergency: t=s+1,s+2,………………………(28) Where t is the time step, the time of occurrence of the emergency is t=s, C_charge represents the rate of change of energy storage, E_supply represents the amount of electricity supplied, and E_demand represents the amount of electricity demanded.

Images