Public building carbon emission prediction method considering multi-factor influence

Abstract

The invention discloses a public building carbon emission prediction method considering multi-factor influence, and belongs to the technical field of energy management. The prediction method comprises the following steps: step 1, acquiring historical hour-level energy consumption data, dynamic carbon emission factor data and human traffic data of a public building, and temperature, humidity and wind speed data of an area where the public building is located; 2, decomposing the historical public building energy consumption data and the dynamic carbon emission factor data into a specified number of intrinsic mode components by using a variational mode decomposition algorithm; step 3, analyzing influence factors of public building energy consumption and dynamic carbon emission factors by using MIC, step 4, optimizing parameters of the HKELM model by using a grey wolf algorithm, and constructing an optimal prediction model; and 5, respectively predicting the eigencomponents by using the trained model to obtain a final public building carbon emission prediction result. According to the invention, the problems of poor pertinence, large prediction data granularity and low prediction precision of carbon emission prediction of public buildings can be effectively solved.

Description

Technical Field

[0001] The present invention relates to the technical field of energy management, and more particularly to a method for predicting carbon emissions from public buildings taking into account the influence of multiple factors. Background Art

[0002] The construction industry is a key industry for energy conservation, emission reduction and achieving the "dual carbon" goals in my country. Its emission reduction potential is huge and the task is arduous. Emission reduction during the building operation stage is a key stage for the construction industry to achieve its emission reduction goals.

[0003] Currently, domestic researchers have conducted research on carbon emission accounting, quantitative diagnosis, and emission reduction for public buildings, but no method for predicting public building carbon emissions has been published. The main research methods for predicting building energy consumption and carbon emissions include the LEAP model, the STIRPAT model, and the system dynamics method. These methods have the following major shortcomings: First, they predict total carbon emissions from a macro perspective, such as the entire construction industry or a specific building category, using national or regional average data to calculate the material or energy consumption of the target building. This does not account for regional variations in building carbon emissions and results in low accuracy and precision. Second, these methods calculate carbon emissions and analyze energy savings over the entire lifecycle of a single building, lacking a more in-depth and accurate carbon emission prediction analysis during the building's operation phase, thus lacking guidance for energy conservation and emission reduction in daily building operations. Third, they fail to conduct building carbon emission prediction and energy conservation analysis for different business types, resulting in insufficiently targeted prediction results and energy conservation analysis, which can significantly differ from actual conditions. Therefore, we propose a method for predicting public building carbon emissions that considers the influence of multiple factors. Summary of the Invention

[0004] The purpose of the present invention is to provide a method for predicting carbon emissions from public buildings that takes into account the influence of multiple factors, so as to solve the problems raised in the above background technology: To achieve the above object, the present invention provides the following technical solutions: A method for predicting carbon emissions from public buildings considering multiple factors includes the following steps: Step 1: Obtain historical hourly energy consumption data, dynamic carbon emission factor data, pedestrian flow data, and temperature, humidity, and wind speed data of public buildings; Step 2: Using the variational mode decomposition (VMD) algorithm, the historical hourly energy consumption data and dynamic carbon emission factor data of public buildings are decomposed into a specified number of eigenmode components according to the preset mode number K and penalty coefficient α; Step 3: Use the Maximum Mutual Information Coefficient (MIC) to analyze the factors influencing the historical hourly energy consumption and dynamic carbon emission factors of public buildings. These factors include temperature, pedestrian flow, humidity, and wind speed. The correlation between each factor and the energy consumption and dynamic carbon emission factors of public buildings is calculated. The top three factors with the highest correlation are selected and used as partial input data for each eigenmode component prediction model. Step 4: Use the Grey Wolf Algorithm (GWO) to optimize the parameters of the HKELM model and build a prediction model. Use the prediction model to construct the GWO-HKELM model for the decomposed intrinsic mode components and train them. Step 5. Use the trained GWO-HKELM model to predict the intrinsic mode components separately, superimpose the public building energy consumption and dynamic carbon emission factor prediction results obtained for each component, and obtain the final public building hourly energy consumption and dynamic carbon emission factor prediction results. Multiply the final public building hourly energy consumption and dynamic carbon emission factor prediction results to obtain the final public building carbon emission prediction results.

[0005] Preferably, step 2 includes the following steps: Step 2.1: Construct the variational optimal problem formula and transform the historical energy consumption data of public buildings into Decompose into K finite bandwidth modal functions ; ; Where K is the number of modes to be decomposed; 、 is the kth modal decomposition and center frequency after decomposition; is an imaginary unit; Indicates the moment; is the time derivative operator (partial derivative with respect to time), which is used to describe the instantaneous frequency characteristics of the signal; * is the convolution operator; is the historical energy consumption data of public buildings; e is the natural logarithm; is the Dirac function; Step 2.2: Use the quadratic penalty factor and Lagrange multiplier to transform the constrained variational problem into an unconstrained variational problem.

[0006] ; Where, is a quadratic penalty factor used to reduce the interference of Gaussian noise; Indicates inner product; Step 2.3: Use the alternating direction multiplier iteration algorithm to solve the unconstrained problem and update iteratively 、 Get the specified number of components; Step 2.4: Process the historical dynamic carbon emission factor data and repeat steps 2.1 to 2.3 to obtain the specified number of components after decomposition of the dynamic carbon emission factor data.

[0007] Preferably, step 4 includes the following steps: Step 4.1: Read the intrinsic component building energy consumption data and construct the HKELM model using the polynomial kernel function, Gaussian function and ELM; Step 4.2, the initial gray wolf algorithm parameters include: population size, number of iterations, and search interval; Step 4.3. Calculate the fitness value of the Grey Wolf Algorithm (GWO) according to the formula: ; Where: D is the number of training samples; is the predicted value; is the true value; Step 4.4: Determine whether the maximum number of iterations has been reached. If so, enter the optimal parameters. Otherwise, return to step 4.3. Step 4.5: Assign the obtained optimal parameters to the HKELM model, train the model, and obtain the optimized model; Step 4.6: Read the intrinsic component building dynamic carbon emission factor data and repeat steps 4.1 to 4.6 to obtain the optimized dynamic carbon emission factor prediction model.

[0008] Preferably, in step 1, the method of obtaining data includes real-time collection through sensors and retrieval from a historical database.

[0009] Preferably, in step 3, the degree of correlation between each influencing factor and the energy consumption and dynamic carbon emission factor of public buildings is calculated, specifically by quantifying the nonlinear correlation between each influencing factor and the energy consumption and dynamic carbon emission factor of public buildings through the maximum mutual information coefficient (MIC) calculation formula.

[0010] Compared with the prior art, the present invention has the following beneficial effects: (1) This invention proposes and implements hourly carbon emission prediction for public buildings. Compared with traditional prediction methods, the prediction object is more targeted and refined, which can closely meet the current demand for building carbon emission supervision in the construction field, provide more timely and accurate data support for building energy consumption management and carbon emission reduction strategy formulation, and help realize the refined management and control of carbon emissions in the construction field.

[0011] (2) This invention combines the historical carbon emission data of public buildings with multiple influencing factors such as pedestrian flow, temperature, humidity, and wind speed, breaking through the limitations of a single data dimension and comprehensively considering the complex factors affecting the carbon emissions of public buildings. Through multi-factor collaborative analysis, it can more realistically reflect the changing patterns of carbon emissions from public buildings, thereby effectively improving the accuracy and precision of the prediction and providing more reliable results for carbon emission prediction.

[0012] (3) The present invention combines the MIC factor analysis algorithm with the Grey Wolf Algorithm (GWO) to improve the HKELM model algorithm. The MIC factor analysis algorithm can scientifically screen out the key factors that have a high degree of influence on the carbon emissions of public buildings and reduce redundant data interference; the Grey Wolf Algorithm (GWO) optimizes the parameters of the HKELM model to enhance the learning and generalization capabilities of the model. The synergistic effect of the two significantly improves the accuracy and efficiency of public building carbon emission prediction, providing technical support for the rapid and accurate prediction of public building carbon emissions. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 It is a schematic diagram of the process of the present invention. DETAILED DESCRIPTION

[0014] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0015] Example: See also Figure 1 , a method for predicting carbon emissions from public buildings considering the influence of multiple factors, including the following steps: Step 1: Obtain historical hourly energy consumption data, dynamic carbon emission factor data, pedestrian flow data, and temperature, humidity, and wind speed data of public buildings. In step 1, data acquisition methods include real-time collection through sensors and retrieval from historical databases.

[0016] Step 2: Using the variational mode decomposition (VMD) algorithm, the historical hourly energy consumption data and dynamic carbon emission factor data of public buildings are decomposed into a specified number of eigenmode components according to the preset mode number K and penalty coefficient α; Specifically, step 2 includes the following steps: Step 2.1: Construct the variational optimal problem formula and transform the historical energy consumption data of public buildings into Decompose into K finite bandwidth modal functions ; ; Where K is the number of modes to be decomposed; 、 is the kth modal decomposition and center frequency after decomposition; is an imaginary unit; Indicates the moment; is the time derivative operator (partial derivative with respect to time), which is used to describe the instantaneous frequency characteristics of the signal; * is the convolution operator; is the historical energy consumption data of public buildings; e is the natural logarithm; is the Dirac delta function, which is used to construct the kernel function in the variational problem; st is the abbreviation of "subject to", which is used to express the constraints of the optimization problem.

[0017] Step 2.2: Use the quadratic penalty factor and Lagrange multiplier to transform the constrained variational problem into an unconstrained variational problem.

[0018] ; Where, is a quadratic penalty factor used to reduce the interference of Gaussian noise; represents the inner product; L( ) represents the augmented Lagrangian function.

[0019] Step 2.3: Use the alternating direction multiplier iteration algorithm to solve the unconstrained problem and update iteratively 、 Get the specified number of components; Step 2.4: Process the historical dynamic carbon emission factor data and repeat steps 2.1 to 2.3 to obtain the specified number of components after decomposition of the dynamic carbon emission factor data.

[0020] Step 3: Use the Maximum Mutual Information Coefficient (MIC) to analyze the factors influencing the historical hourly energy consumption and dynamic carbon emission factors of public buildings. The influencing factors include temperature, pedestrian flow, humidity, wind speed, etc. The correlation between each influencing factor and the public building energy consumption and dynamic carbon emission factors is calculated. The data of the top three influencing factors with the highest correlation are selected as part of the input data for each eigenmode component prediction model. Specifically, the correlation between each influencing factor and the public building energy consumption and dynamic carbon emission factors is calculated by quantifying the nonlinear correlation between each influencing factor and the public building energy consumption and dynamic carbon emission factors through the maximum mutual information coefficient (MIC) calculation formula.

[0021] Step 4: Use the Grey Wolf Algorithm (GWO) to optimize the parameters of the HKELM model and build a prediction model. Use the prediction model to construct the GWO-HKELM model for the decomposed intrinsic mode components and train them. Specifically, step 4 includes the following steps: Step 4.1: Read the intrinsic component building energy consumption data and construct the HKELM model using the polynomial kernel function, Gaussian function and ELM; Step 4.2: Initial gray wolf algorithm parameters include: population size, number of iterations, search interval, etc. Step 4.3. Calculate the fitness value of the Grey Wolf Algorithm (GWO) according to the formula: ; Where: D is the number of training samples; is the predicted value; is the true value; Step 4.4: Determine whether the maximum number of iterations has been reached. If so, enter the optimal parameters. Otherwise, return to step 4.3. Step 4.5: Assign the obtained optimal parameters to the HKELM model, train the model, and obtain the optimized model; Step 4.6: Read the intrinsic component building dynamic carbon emission factor data and repeat steps 4.1 to 4.6 to obtain the optimized dynamic carbon emission factor prediction model.

[0022] Step 5. Use the trained GWO-HKELM model to predict the intrinsic mode components separately. Specifically, input the test data and use the trained GWO-HKELM model of each component to predict the component energy consumption and dynamic carbon emission factor; superimpose the public building energy consumption and dynamic carbon emission factor prediction results obtained for each component to obtain the final public building hourly energy consumption and dynamic carbon emission factor prediction results; multiply the final public building hourly energy consumption and dynamic carbon emission factor prediction results to obtain the final public building carbon emission prediction results.

[0023] The present invention can effectively solve the problems of the lack of pertinence in public building carbon emission prediction, large granularity of prediction data, and low prediction accuracy through the above-mentioned method steps. The present invention proposes and implements hourly carbon emission prediction for public buildings. Compared with traditional prediction methods, the prediction object is more targeted and refined, which can closely meet the current demand for building carbon emission supervision in the construction field, provide more timely and accurate data support for building energy consumption management and carbon emission reduction strategy formulation, and help realize the refined management and control of carbon emissions in the construction field. Combining the historical carbon emission data of public buildings with multiple influencing factors such as pedestrian flow, temperature, humidity, wind speed, etc., breaks through the limitations of a single data dimension, comprehensively considers the complex factors affecting the carbon emissions of public buildings, and through multi-factor collaborative analysis, can more realistically reflect the changing laws of public building carbon emissions, thereby effectively improving the accuracy and precision of the prediction, and providing more reliable results for carbon emission prediction. The method of combining the MIC factor analysis algorithm with the Grey Wolf Algorithm (GWO) to improve the HKELM model algorithm is used. The MIC factor analysis algorithm can scientifically screen out key factors that have a high degree of impact on carbon emissions from public buildings and reduce redundant data interference; the Grey Wolf Algorithm (GWO) optimizes the parameters of the HKELM model to enhance the model's learning and generalization capabilities. The synergistic effect of the two significantly improves the accuracy and efficiency of public building carbon emission prediction, providing technical support for the rapid and accurate prediction of public building carbon emissions.

[0024] The above shows and describes the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions are merely preferred examples of the present invention and are not intended to limit the present invention. Various changes and improvements may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and improvements fall within the scope of the present invention. The scope of protection claimed in the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for predicting carbon emissions from public buildings considering multiple factors, characterized in that: The steps include: Step 1: Obtain historical hourly energy consumption data, dynamic carbon emission factor data, pedestrian flow data, and temperature, humidity, and wind speed data of public buildings; Step 2: Using the variational mode decomposition (VMD) algorithm, the historical hourly energy consumption data and dynamic carbon emission factor data of the public building are decomposed into a specified number of eigenmode components according to a preset mode number K and penalty coefficient α; Step 3: Use the Maximum Mutual Information Coefficient (MIC) to analyze the factors influencing the historical hourly energy consumption and dynamic carbon emission factors of public buildings. The factors include temperature, pedestrian flow, humidity, and wind speed. The correlation between each factor and the energy consumption and dynamic carbon emission factors of public buildings is calculated. The top three factors with the highest correlation are selected and used as partial input data for each eigenmode component prediction model. Step 4: Use the Grey Wolf Algorithm (GWO) to optimize the parameters of the HKELM model and build a prediction model. Use the prediction model to construct the GWO-HKELM model for the decomposed intrinsic mode components and train them. Step 5. Use the trained GWO-HKELM model to predict the intrinsic mode components separately, superimpose the public building energy consumption and dynamic carbon emission factor prediction results obtained for each component, and obtain the final public building hourly energy consumption and dynamic carbon emission factor prediction results. Multiply the final public building hourly energy consumption and dynamic carbon emission factor prediction results to obtain the final public building carbon emission prediction results.

2. The method for predicting carbon emissions from public buildings considering multiple factors according to claim 1, characterized in that: The step 2 comprises the following steps: Step 2.1: Construct the variational optimal problem formula and transform the historical energy consumption data of public buildings into Decompose into K finite bandwidth modal functions ; Where K is the number of modes to be decomposed; 、 is the kth modal decomposition and center frequency after decomposition; is an imaginary unit; Indicates the moment; is the time derivative operator, used to describe the instantaneous frequency characteristics of the signal; * is the convolution operator; is the historical energy consumption data of public buildings; e is the natural logarithm; is the Dirac function; Step 2.2: Use the quadratic penalty factor and Lagrange multiplier to transform the constrained variational problem into an unconstrained variational problem. Where, is a quadratic penalty factor used to reduce the interference of Gaussian noise; Indicates inner product; Step 2.3: Use the alternating direction multiplier iteration algorithm to solve the unconstrained problem and update iteratively 、 Get the specified number of components; Step 2.4: Process the historical dynamic carbon emission factor data and repeat steps 2.1 to 2.3 to obtain the specified number of components after decomposition of the dynamic carbon emission factor data.

3. The method for predicting carbon emissions from public buildings considering multiple factors according to claim 1, characterized in that: The step 4 comprises the following steps: Step 4.1: Read the intrinsic component building energy consumption data and construct the HKELM model using the polynomial kernel function, Gaussian function and ELM; Step 4.2, the initial gray wolf algorithm parameters include: population size, number of iterations, and search interval; Step 4.

3. Calculate the fitness value of the Grey Wolf Algorithm (GWO) according to the formula: Where: D is the number of training samples; is the predicted value; is the true value; Step 4.4: Determine whether the maximum number of iterations has been reached. If so, enter the optimal parameters. Otherwise, return to step 4.

3. Step 4.5: Assign the obtained optimal parameters to the HKELM model, train the model, and obtain the optimized model; Step 4.6: Read the intrinsic component building dynamic carbon emission factor data and repeat steps 4.1 to 4.6 to obtain the optimized dynamic carbon emission factor prediction model.

4. The method for predicting carbon emissions from public buildings considering multiple factors according to claim 1 is characterized by: In step 1, the data is obtained by collecting data in real time through sensors and retrieving data from a historical database.

5. The method for predicting carbon emissions from public buildings considering multiple factors according to claim 1 is characterized by: In step 3, the degree of correlation between each influencing factor and the energy consumption and dynamic carbon emission factor of public buildings is calculated. Specifically, the nonlinear correlation between each influencing factor and the energy consumption and dynamic carbon emission factor of public buildings is quantified through the maximum mutual information coefficient (MIC) calculation formula.

Images