Parameterization characterization method for quantizing prediction precision and difficulty of building energy consumption curve

Abstract

The invention provides a parameterized characterization method for quantifying building energy consumption curve prediction precision and difficulty. The method comprises the following steps: dividing a building group according to types; building a building type consistency index, wherein the building type consistency index is used for measuring the similarity of the same type of buildings in the aspect of power consumption behaviors; obtaining building type consistency indexes of building groups of the same type; constructing an electrical load prediction model of the building group of the same type; inputting the historical power utilization data set of the building groups of the same type into the power utilization load prediction model of the building groups of the same type to obtain the prediction precision of the power utilization load prediction model of the building groups of the same type; and establishing a first mapping relationship between the building type consistency index of the building groups of the same type and the prediction precision of the electrical load prediction model of the building groups of the same type. The method can assist in completing data screening of tasks in energy consumption prediction of large-scale building groups.

Description

Technical Field

[0001] This application relates to the field of building energy consumption prediction technology, and in particular to a parametric characterization method for quantifying the accuracy and difficulty of predicting building energy consumption curves. Background Technology

[0002] The deep integration of buildings and power systems has become an important direction for energy transition. New power systems are facing output volatility and uncertainty due to both high winter and summer loads and the increasing penetration of renewable energy. This challenge is particularly pronounced on the building side. With the continuous increase in distributed photovoltaic capacity and the gradual application of energy storage technology, building energy loads exhibit greater dynamism and uncertainty, further exacerbating the regulation pressure on the power system. Against this backdrop, building load forecasting and energy optimization have become crucial means to achieve source-load synergy, alleviate power peak-shaving pressure, and improve system stability. In particular, accurate forecasting and characterization of power load are key foundations for achieving building energy efficiency management, proactive load response, and flexible power system regulation. With the continuous improvement of smart meters and the digitalization of the power grid, building energy consumption data is exploding, providing rich data support for building high-precision load forecasting models. Current mainstream building load forecasting methods are mostly based on a combination of historical data and machine learning algorithms, evolving from early traditional neural network models to Long Short-Term Memory (LS1TM) networks, attention mechanism networks, and the recently emerging methods such as active learning and deep reinforcement learning. These methods, while improving prediction accuracy, are gradually showing a trend of reducing data dependence and improving data utilization efficiency.

[0003] However, existing technologies primarily focus on improving predictive performance from the perspective of data structure or data selection. For example, they employ preprocessing techniques such as Variational Mode Decomposition (VMD) or cluster analysis to remove noise and unpredictable factors; or they use Principal Component Analysis (PCA) for dimensionality reduction and active learning mechanisms to select data with valuable information for model training. These strategies are effective for predictive tasks in small- to medium-sized buildings or specific areas, but when faced with massive energy consumption data generated by regional or even city-level building clusters, the computational complexity, model adaptability, and data processing costs of these algorithms increase significantly, severely limiting their practicality and widespread applicability. Summary of the Invention

[0004] The purpose of this application is to provide a parameterized characterization method for quantifying the accuracy and difficulty of building energy consumption curve prediction, which can assist in data screening for tasks in large-scale building group energy consumption prediction.

[0005] One aspect of this application provides a parametric characterization method for quantifying the accuracy and difficulty of predicting building energy consumption curves. The method includes: classifying building groups according to type; constructing a building type consistency index, which measures the similarity of electricity consumption behavior among buildings of the same type; obtaining the building type consistency index for building groups of the same type; constructing an electricity load prediction model for the building groups of the same type; inputting historical electricity consumption datasets of the building groups of the same type into the electricity load prediction model for the building groups of the same type to obtain the prediction accuracy of the electricity load prediction model for the building groups of the same type; and establishing a first mapping relationship between the building type consistency index of the building groups of the same type and the prediction accuracy of the electricity load prediction model for the building groups of the same type.

[0006] Furthermore, the construction of the building type consistency index includes: calculating the correlation between the electricity load curves of each pair of buildings in a building group of the same type to obtain a correlation coefficient sample set of the building group of the same type; fitting the probability distribution of the correlation coefficient sample set to select the optimal probability distribution; extracting the parameters of the optimal probability distribution, and using the parameters of the optimal probability distribution as the building type consistency index.

[0007] Furthermore, the correlation calculation of the electricity load curves between any two buildings in a building group of the same type includes: using the Pearson correlation coefficient to calculate the correlation between the electricity load curves between any two buildings in a building group of the same type.

[0008] Further, the step of fitting the probability distribution of the correlation coefficient sample set and selecting the optimal probability distribution includes: fitting the probability distribution of the correlation coefficient sample set using multiple candidate probability distributions respectively; evaluating the goodness of fit of the multiple candidate probability distributions using the Akaike information criterion respectively; and selecting the optimal probability distribution from the multiple candidate probability distributions.

[0009] Further, the step of fitting the correlation coefficient sample set with multiple candidate probability distributions includes fitting the correlation coefficient sample set with a normal distribution, a gamma distribution, and a Beta distribution, respectively; the step of selecting the optimal probability distribution from the multiple candidate probability distributions includes selecting the probability distribution with the smallest AIC from the normal distribution, the gamma distribution, and the Beta distribution as the optimal probability distribution, wherein the optimal probability distribution includes the Beta distribution.

[0010] Furthermore, the method further includes: extracting an orderliness index for an individual building, wherein the orderliness index is used to reflect the periodicity and regularity of the electricity load curve of the individual building in the time domain; constructing an electricity load prediction model for the individual building; inputting the historical electricity consumption dataset of the individual building into the electricity load prediction model for the individual building to obtain the prediction accuracy of the electricity load prediction model for the individual building; and establishing a second mapping relationship between the orderliness index of the individual building and the prediction accuracy of the electricity load prediction model for the individual building.

[0011] Furthermore, the extraction of the orderliness index of individual buildings includes: performing time-domain to frequency-domain conversion on the electricity load curve of the individual buildings to obtain an electricity load spectrum diagram; extracting the amplitude of the main frequency components in a predetermined period of the electricity load spectrum diagram; calculating the ratio of the sum of the amplitudes of the main frequency components in the predetermined period to the total spectral amplitude of the electricity load spectrum diagram, and using the ratio as the orderliness index.

[0012] Furthermore, the method also includes: classifying building groups and individual buildings according to prediction difficulty based on the first mapping relationship and the second mapping relationship; and adopting corresponding electricity consumption data filtering strategies based on the prediction difficulty classification.

[0013] Furthermore, the classification of building groups and individual buildings according to prediction difficulty includes: dividing building groups and individual buildings into three categories: easy to predict, medium to difficult to predict, and difficult to predict; the adoption of corresponding data filtering strategies based on prediction difficulty classification includes: for easy-to-predict buildings, using a unified electricity load prediction model to perform batch electricity load prediction; for medium-difficulty buildings, further optimizing the structure of the electricity load prediction model; and for difficult-to-predict buildings, supplementing the sampling frequency, improving data quality, or constructing a personalized electricity load prediction model.

[0014] Furthermore, the method also includes assigning prediction difficulty level labels to all buildings for easy subsequent invocation.

[0015] The parameterized characterization method for the prediction accuracy and difficulty of quantitative building energy consumption curves proposed in this application constructs a building type consistency index and establishes a first mapping relationship between the building type consistency index and the prediction accuracy of its electricity load prediction model. This can form a feasibility evaluation system for overall building prediction, which can be used to quickly assess the load prediction potential of any group of building types.

[0016] The parameterized characterization method for predicting the accuracy and difficulty of building energy consumption curves proposed in this application constructs an orderliness index for individual buildings and establishes a second mapping relationship between the orderliness index of individual buildings and the prediction accuracy of their electricity load prediction models. This method can be used to quickly determine the data quality and modeling difficulty at the individual building level, and assist in model design and scheduling strategy formulation.

[0017] The parameterized characterization method for quantifying the prediction accuracy and difficulty of building energy consumption curves in this application can quickly assess the predictability of building energy consumption data in load forecasting tasks without the need for complex modeling or deep learning algorithms. It clarifies the data's utilization value and modeling difficulty, thereby enabling rapid screening and classification of high-quality load data—that is, quantifying its prediction accuracy and difficulty. This helps to quickly identify key datasets with high information quality and strong model adaptability from massive amounts of data. This not only helps improve the overall efficiency, applicability, and scalability of subsequent building energy consumption prediction models but also provides support for higher-level building energy consumption management and power system dispatch optimization. Attached Figure Description

[0018] Figure 1 This is a flowchart illustrating a parametric characterization method for predicting the accuracy and difficulty of building energy consumption curves according to an embodiment of this application.

[0019] Figure 2 Another flowchart of a parameterized characterization method for predicting the accuracy and difficulty of building energy consumption curves according to an embodiment of this application.

[0020] Figure 3 A comparative diagram of AICs for different building types.

[0021] Figure 4 The figure shows the fitting effect of the correlation coefficient distribution of the five types of buildings.

[0022] Figure 5 This represents the mapping relationship between building type consistency and prediction accuracy.

[0023] Figure 6 This represents the mapping relationship between the average correlation coefficient and the prediction accuracy.

[0024] Figure 7 A schematic diagram illustrating the extraction and parametric representation of orderliness indicators for individual buildings.

[0025] Figure 8 This represents the mapping relationship between the orderliness index of office buildings and the prediction accuracy.

[0026] Figure 9 This is a schematic diagram illustrating the prediction difficulty level classification and data filtering strategy of this application. Detailed Implementation

[0027] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numerals in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses consistent with some aspects of this application as detailed in the appended claims.

[0028] This application provides a parametric characterization method for quantifying the accuracy and difficulty of building energy consumption curve prediction. It can quickly assess the predictability of building energy consumption data in load forecasting tasks without the need for complex modeling or deep learning algorithms, clarifying the data's utilization value and modeling difficulty. This enables rapid screening and classification of high-quality load data, i.e., quantifying its prediction accuracy and difficulty, to assist in quickly identifying key datasets with high information quality and strong model adaptability from massive datasets. This not only helps improve the overall efficiency, applicability, and scalability of subsequent building energy consumption prediction models but also provides support for higher-level building energy consumption management and power system dispatch optimization.

[0029] The parametric characterization method for predicting the accuracy and difficulty of quantitative building energy consumption curves according to this application will be described in detail below with reference to the accompanying drawings. Unless otherwise specified, the features in the following embodiments and implementation methods can be combined with each other.

[0030] Figure 1 A flowchart illustrating a parametric characterization method for predicting the accuracy and difficulty of building energy consumption curves according to an embodiment of this application is provided. Figure 1 As shown, a parametric characterization method for predicting the accuracy and difficulty of building energy consumption curves according to an embodiment of this application may include steps S11 to S16.

[0031] In step S11, the building groups are divided according to type.

[0032] For example, based on type, building complexes can be divided into offices, hospitals, shopping malls, schools, hotels, etc.

[0033] In step S12, a building type consistency index is constructed, which is used to measure the similarity of electricity consumption behavior among buildings of the same type.

[0034] Buildings with the same purpose or function often exhibit similar daily electricity consumption patterns, which is crucial for constructing a unified prediction model. Therefore, this application uses a building type consistency index to measure the degree of consistency in electricity consumption behavior among buildings of the same type.

[0035] In some embodiments, the building type consistency index in step S12 may further include steps S121 to S123.

[0036] In step S121, the correlation between the electricity load curves of each pair of buildings in the same type of building group is calculated to obtain the correlation coefficient sample set of the same type of building group.

[0037] Optionally, the Pearson correlation coefficient can be used to calculate the correlation between the electricity load curves of pairs of buildings within a building group of the same type. The closer the correlation coefficient is to 1, the more consistent the electricity load patterns among the buildings, which is more conducive to building a shared model. For buildings of the same type, a correlation coefficient close to 1 is generally desirable. Conversely, if the correlation coefficient is close to 0 or negative, it indicates significant differences in electricity consumption behavior among buildings, resulting in poor adaptability of the general model. By constructing a correlation coefficient matrix for a building group, the overall consistency of load behavior among buildings of the same type can be preliminarily characterized.

[0038] The formula for calculating the Pearson correlation coefficient is shown below: (1) In the formula, This represents the Pearson correlation coefficient. and This represents the electrical load of two similar buildings at the i-th sampling time. and This represents the average electrical load of two buildings of the same type.

[0039] By using the Pearson correlation coefficient to calculate the correlation between each pair of buildings in a group of buildings of the same type, the correlation coefficient between each pair of buildings can be obtained, thus ultimately yielding a sample set of correlation coefficients for the group of buildings of the same type.

[0040] In step S122, the probability distribution of the correlation coefficient sample set is fitted, and the optimal probability distribution is selected.

[0041] In some embodiments, step S122, which involves fitting a probability distribution to the correlation coefficient sample set and selecting the optimal probability distribution, may further include steps S1221 and S1222.

[0042] In step S1221, multiple candidate probability distributions are used to fit the probability distribution of the correlation coefficient sample set.

[0043] For example, multiple candidate probability distributions such as normal distribution, gamma distribution, and beta distribution can be used to fit the probability distribution of the correlation coefficient sample set.

[0044] In step S1222, the goodness of fit of multiple candidate probability distributions is evaluated using the Akaike Information Criterion (AIC), and the optimal probability distribution is selected from the multiple candidate probability distributions.

[0045] For example, the Akaike information criterion can be used to select the probability distribution with the minimum AIC from multiple candidate probability distributions such as the normal distribution, gamma distribution, and beta distribution as the optimal probability distribution. In this application, the beta distribution was ultimately selected as the optimal probability distribution.

[0046] The formula for calculating the Akaike information criterion is as follows: (2) Where AIC represents the Akaike information criterion; k represents the number of free parameters (parameters to be estimated) in the model; and L represents the likelihood function.

[0047] In step S123, the parameters of the optimal probability distribution are extracted and used as the building type consistency index.

[0048] This application ultimately selected the Beta distribution as the optimal probability distribution, using it as the core representation method for the building type consistency index, and extracting its shape parameters α and β as building type consistency indicators. The shape parameters of the Beta distribution can flexibly describe the central tendency and skewness of the data, which gives it a significant advantage in characterizing the consistency of building electricity consumption.

[0049] The probability density function of the Beta distribution is shown below: (3) in, The probability density function (PDF) represents the Beta distribution. This represents the Beta function, used for normalization, ensuring that the integral of the probability density over [0,1] is 1; This represents the normalized sample data (i.e., the Pearson correlation coefficient between buildings in this application). ); and Let be the two shape parameters of the Beta distribution.

[0050] (4) in, Indicates the integral variable (without practical meaning); This represents the Gamma function.

[0051] In step S13, the building type consistency index of the same type of building group is obtained.

[0052] In step S14, a unified electricity load prediction model for the same type of building group is constructed, such as one based on XGBoost (eXtreme Gradient Boosting) or LSTM (Long Short-Term Memory).

[0053] In step S15, the historical electricity consumption dataset of the same type of building group is input into the electricity load prediction model of the same type of building group to obtain the prediction accuracy of the electricity load prediction model of the same type of building group, such as RMSE (Root Mean Squared Error) and MAPE (Mean Absolute Percentage Error).

[0054] In step S16, a first mapping relationship is established between the building type consistency index of the same type of building group and the prediction accuracy of the electricity load prediction model of the same type of building group.

[0055] The parametric characterization method of this application constructs a building type consistency index and establishes a first mapping relationship between the building type consistency index of the same type of building group and the prediction accuracy of its electricity load prediction model. This can form a feasibility evaluation system for overall building prediction. The first mapping relationship can be used to quickly assess the load prediction potential of any building type group.

[0056] The above addresses the level of building groups. At the individual building level, to assess the difficulty of predicting data for individual buildings, Figure 2 Another flowchart of a parameterized characterization method for predicting the accuracy and difficulty of building energy consumption curves, according to an embodiment of this application, is disclosed. Figure 2 As shown, in some embodiments, the parameterization characterization method of this application may further include steps S21 to S24.

[0057] In step S21, the orderliness index of individual buildings is extracted. The orderliness index is used to reflect the periodicity and regularity of the electricity load curve of individual buildings in the time domain.

[0058] In some embodiments, step S21, which extracts the orderliness index of individual buildings, may further include steps S211 to S213.

[0059] In step S211, Fourier transform is applied to the electrical load curve of each building to convert it from the time domain to the frequency domain, so as to obtain the electrical load spectrum diagram.

[0060] In step S212, the amplitude of the main frequency component on a predetermined period of the electrical load spectrum is extracted.

[0061] For example, the amplitude of the main frequency components in a specific period (such as 12 hours, 24 hours, 168 hours) of the electricity load spectrum can be extracted.

[0062] In step S213, the ratio of the sum of the amplitudes of the main frequency components on the predetermined period to the total spectral amplitude of the power load spectrum is calculated, and the ratio is used as an orderliness index.

[0063] A higher orderliness index indicates a more regular and periodic electricity load curve, which is beneficial for model learning and prediction. The orderliness index has unit normalization properties and can be compared across buildings.

[0064] The orderliness index can be represented as follows: (5) Here, OC represents the orderliness index. , , These represent the amplitudes of the principal frequency components over 12 hours, 24 hours, and 168 hours, respectively. This represents the amplitude at the corresponding frequency.

[0065] (6) in, This represents the frequency domain signal after the discrete Fourier transform (the complex representation of the k-th frequency component). This represents the value of the t-th sample point of the original discrete signal (time-domain signal).

[0066] (7) in, This represents the amplitude of a frequency component, corresponding to the k-th frequency component. Representing complex numbers The model; Representing complex numbers The real part; Representing complex numbers The imaginary part.

[0067] Continue to refer to Figure 2 In step S22, a predictive model for the electricity load of individual buildings is constructed.

[0068] For example, lightweight models such as SVR (Support Vector Regression) and KNN (K-Nearest Neighbors) can be used.

[0069] In step S23, the historical electricity consumption data of individual buildings is input into the electricity load prediction model of individual buildings to obtain the prediction accuracy of the electricity load prediction model of individual buildings.

[0070] In step S24, a second mapping relationship is established between the orderliness index of individual buildings and the prediction accuracy of the electricity load prediction model of individual buildings.

[0071] The parametric characterization method of this application constructs an orderliness index for individual buildings and establishes a second mapping relationship between the orderliness index of individual buildings and the prediction accuracy of their electricity load prediction models. This can be used to quickly determine the data quality and modeling difficulty at the individual building level, and assist in model design and scheduling strategy formulation.

[0072] In some embodiments, the parameterization characterization method of this application may further include steps S31 and S32.

[0073] In step S31, based on the first mapping relationship and the second mapping relationship, the building groups and individual buildings are classified according to the prediction difficulty.

[0074] For example, building groups and individual buildings can be divided into three categories according to the difficulty of prediction: easy to predict (i.e., low difficulty prediction), medium difficulty prediction, and difficult to predict (i.e., high difficulty prediction).

[0075] In step S32, a corresponding electricity consumption data filtering strategy is adopted based on the prediction difficulty level.

[0076] For example, for buildings with easily predictable electricity loads, a unified electricity load forecasting model can be used for batch electricity load forecasting; for buildings with moderately difficult forecasting, the structure of the electricity load forecasting model can be further optimized; for buildings with difficult forecasting, the sampling frequency can be increased, data quality can be improved, or a personalized electricity load forecasting model can be built. Finally, the data is prioritized based on the indicators, which serves as the basis for selecting data for subsequent feature engineering and model training.

[0077] In some embodiments, the parameterization characterization method of this application may further include step S33.

[0078] In step S33, all buildings are assigned a prediction difficulty level label for easy retrieval later.

[0079] Compared with existing technologies, the parametric characterization method for predicting the accuracy and difficulty of quantitative building energy consumption curves in this application has the following advantages: Traditional building load forecasting methods generally rely on deep learning or complex feature engineering. Model training is highly dependent on data volume and computational resources, making rapid deployment across city-level building clusters difficult. In contrast, the parametric characterization method for quantifying the accuracy and difficulty of building energy consumption curve prediction proposed in this application eliminates the need for complex models. It allows for data quality assessment and difficulty discrimination before prediction, significantly improving the overall system response speed and applicability. This application constructs building type consistency indices for building clusters and orderliness indices for individual buildings through correlation analysis between building types and frequency domain orderliness analysis of individual buildings. Furthermore, it establishes a quantifiable and interpretable prediction difficulty evaluation system by mapping probabilistic model fitting to prediction accuracy. Compared to error outputs relying on black-box models, this parametric characterization method clearly explains the sources of prediction difficulties, facilitating model optimization and data governance. It is suitable for comparative analysis across different regions and building types, enabling horizontal evaluation and prioritization across large-scale building clusters at the regional and city levels. This parametric characterization method helps power load forecasting platforms prioritize high-value data under limited resources, improving overall modeling efficiency and accuracy. With the widespread adoption of smart meters, energy consumption sensors, and building energy management systems, building load data will experience explosive growth. The parameterized characterization method proposed in this application can adapt to this trend, providing a scalable and efficient tool for electricity data governance and evaluation, laying the foundation for building a large-scale, intelligent load forecasting platform in the future.

[0080] The following section will describe in detail the parametric characterization method for predicting the accuracy and difficulty of quantitative building energy consumption curves in this application, using illustrations and specific examples.

[0081] Figure 3 This diagram illustrates a comparison of AICs (Automatic Enclosures) across different building types. Figure 3 As can be seen, compared to the normal and Gamma distributions, the Beta distribution has a smaller AIC index, better reflecting the distribution characteristics of correlation coefficients among similar buildings. Therefore, this application ultimately selected the Beta distribution as the optimal probability distribution.

[0082] Figure 4 The distribution and fitting effect of the correlation coefficients of the five types of buildings are revealed. Table 1 below provides an example of the consistency index of building types for the five building groups.

[0083] Table 1

[0084] like Figure 4Based on Table 1, a Beta distribution was fitted to the correlation coefficients of five typical building types. The shape parameters obtained from the fitting allow for a direct assessment of the consistency of electricity load data across different building types. A larger α value indicates a more rightward sloping distribution curve, suggesting a stronger correlation between buildings of the same type; a larger β value indicates a more leftward sloping distribution curve, suggesting a weaker correlation between buildings of the same type. In the example, the α values ​​for office, hospital, shopping mall, and school buildings are significantly larger than their β values, with the correlation coefficients concentrated on the right side of the distribution curve, i.e., in the high-value range.

[0085] according to Figure 4 The correlation coefficient distribution shown in the figure is in Figure 5 The mapping relationship between building type consistency and prediction accuracy was constructed in the middle. Figure 6 The paper establishes a mapping relationship between the average correlation coefficient and prediction accuracy. In the example, a unified XGBoost model was trained for each type of building to predict electricity load, thereby deriving the average R-value reflecting the prediction accuracy of each type of building. 2 . Figure 5 The x and y axes represent the shape parameters of the Beta distribution, and the point size reflects the prediction accuracy for the corresponding building type. Type consistency based on shape parameters reflects the positive correlation between the correlation coefficient distribution and prediction accuracy. When α is greater than β, the correlation coefficient distribution among the corresponding building types is closer to a strong correlation; when predicting the load of this type of building, machine learning models are more likely to capture consistent electricity load change trends and characteristics, thus achieving higher prediction accuracy. Figure 6 The diagram further illustrates that the higher the correlation coefficient of the electricity load of individual buildings of the same type, the higher the prediction accuracy of that type of building. Based on the mapping relationship between type consistency index and prediction accuracy, model building strategies can be guided in large-scale building cluster energy consumption prediction tasks. For building types with high consistency, a unified prediction model can be used, while for building types with high degree of individuality, a more detailed model can be used to learn the characteristic differences between buildings or to adapt to different buildings by adjusting model parameters.

[0086] Based on the parameterization of prediction accuracy at the building type level, the predictability of individual buildings can be assessed through individual orderliness indicators. Figure 7 This paper presents a schematic diagram illustrating the extraction and parameterized characterization of the orderliness index. For the original building electricity load time series, the time-domain data is transformed to the frequency domain using Fourier transform. Besides the low-frequency, long-period trends, the short-term variations in building electricity consumption within 12 hours, 24 hours (day), and 168 hours (week) better reflect its regularity. Therefore, the amplitudes of the corresponding three periods are found in the frequency domain, and the ratio between these amplitudes and the total spectral amplitude is calculated as the orderliness index. A higher orderliness index indicates that the electricity load curve has a high amplitude at the corresponding frequency, exhibiting obvious periodic characteristics. Figure 7 The figure shows typical weekly electricity consumption curves for buildings with orderliness indices of 0.101 and 0.051. As can be seen from the figure, the higher the orderliness index, the more obvious the periodicity of the electricity load curve; while when the orderliness index is low, the electricity load curve shows a more disordered fluctuation state.

[0087] Figure 7 The electricity consumption regularity index displayed in the paper represents a pattern that can further measure and explain the difficulty of predicting and generating building energy consumption. Theoretically, the higher the regularity index of an individual building, the easier it is to predict and generate it through data-driven or physical modeling methods.

[0088] Figure 8 This study reveals the mapping relationship between the orderliness index and prediction accuracy of office buildings. The orderliness index of individual buildings shows a strong positive correlation with prediction accuracy, exhibiting a linear increasing relationship. This mapping relationship and the orderliness index can be used to quickly determine the predictability of the corresponding building's electrical load.

[0089] Figure 9 This paper illustrates the prediction difficulty level classification and data filtering strategy of this application. The prediction difficulty level classification and data filtering strategy is based on the building type consistency index and building individual orderliness index constructed above, and their mapping relationship with prediction accuracy, further realizing the effective classification and priority filtering of building load data. The specific implementation method is as follows: S81. Input data processing: Obtain historical load data of the building complex and complete the preliminary classification based on the building's use and functional attributes.

[0090] S82. Calculation of Building Type Consistency Index and Orderliness Index: Calculate the Pearson correlation coefficient for the load curves of buildings of the same type, and perform Beta distribution fitting to extract shape parameters (i.e., building type consistency index); perform Fourier transform on each individual building to calculate the amplitude ratio in the 12h, 24h, and 168h frequency bands, and extract the orderliness index.

[0091] S83. Establishment of Prediction Accuracy Estimation Model: Construct an electricity load prediction model using partial data, and establish the mapping relationship between the building type consistency index and the prediction accuracy of the electricity load prediction model, and between the orderliness index and the prediction accuracy of the electricity load prediction model.

[0092] S84. Prediction Difficulty Level Classification: Based on the mapping relationship, the building data is divided into three categories: low difficulty, medium difficulty, and high difficulty. The classification criteria can be set according to the actual accuracy threshold, such as MAPE<10%, 10%~20%, and >20% corresponding to the three levels respectively.

[0093] S85. Data Filtering and Usage Strategy Formulation: S851: Low-difficulty building data is incorporated into a unified prediction model as the main training data; S852: Medium-difficulty data is used for auxiliary purposes such as model optimization, validation, and transfer training; S853: High-difficulty data is incorporated into a separate processing flow, with options for customized modeling, data augmentation, or temporary non-use.

[0094] S86. Create a difficulty level label library: Assign prediction difficulty labels to all building samples for subsequent system calls such as model scheduling, energy consumption management and data governance.

[0095] The parametric characterization method for predicting the accuracy and difficulty of quantitative building energy consumption curves provided in the embodiments of this application has been described in detail above. Specific examples have been used to illustrate the parametric characterization method for predicting the accuracy and difficulty of quantitative building energy consumption curves in the embodiments of this application. The descriptions of the above embodiments are only for helping to understand the core ideas of this application and are not intended to limit this application. It should be noted that for those skilled in the art, several improvements and modifications can be made to this application without departing from the spirit and principles of this application, and these improvements and modifications should all fall within the protection scope of the appended claims.

Claims

1. A parametric characterization method for quantifying the accuracy and difficulty of predicting building energy consumption curves, characterized in that, include: The building complexes are divided according to type; A building type consistency index is constructed, which is used to measure the similarity of electricity consumption behavior among buildings of the same type; Obtain building type consistency indicators for similar building groups; Construct a power load prediction model for this type of building complex; Input the historical electricity consumption dataset of the same type of building group into the electricity load prediction model of the same type of building group to obtain the prediction accuracy of the electricity load prediction model of the same type of building group; Establish a first mapping relationship between the building type consistency index of the same type of building group and the prediction accuracy of the electricity load prediction model of the same type of building group.

2. The parameterized characterization method as described in claim 1, characterized in that, The consistency indicators for the constructed building types include: Correlation calculations are performed on the electricity load curves of each pair of buildings in a building group of the same type to obtain a sample set of correlation coefficients for the building group of the same type. Fit the probability distribution of the correlation coefficient sample set and select the optimal probability distribution; Extract the parameters of the optimal probability distribution and use the parameters of the optimal probability distribution as the consistency index of the building type.

3. The parameterized characterization method as described in claim 2, characterized in that, The correlation calculation of the electricity load curves between any two buildings in a building group of the same type includes: The correlation between the electricity load curves of two buildings in the same building group was calculated using the Pearson correlation coefficient.

4. The parameterized characterization method as described in claim 2, characterized in that, The step of fitting a probability distribution to the correlation coefficient sample set and selecting the optimal probability distribution includes: The correlation coefficient sample set was fitted with probability distributions using multiple candidate probability distributions respectively; The goodness of fit of the multiple candidate probability distributions is evaluated using the Akaike information criterion, and the optimal probability distribution is selected from the multiple candidate probability distributions.

5. The parameterized characterization method as described in claim 4, characterized in that, The step of fitting the correlation coefficient sample set with multiple candidate probability distributions includes: The probability distribution of the correlation coefficient sample set was fitted using normal, gamma, and beta distributions, respectively. The step of selecting the optimal probability distribution from the plurality of candidate probability distributions includes: The probability distribution with the smallest AIC is selected from the normal distribution, the gamma distribution, and the Beta distribution as the optimal probability distribution, wherein the optimal probability distribution includes the Beta distribution.

6. The parameterized characterization method as described in claim 1, characterized in that, Also includes: Extract the orderliness index of individual buildings, which is used to reflect the periodicity and regularity of the electricity load curve of the individual buildings in the time domain; Construct a power load prediction model for the individual buildings; The historical electricity consumption data of the building is input into the electricity load prediction model of the building to obtain the prediction accuracy of the electricity load prediction model of the building. A second mapping relationship is established between the orderliness index of the building individual and the prediction accuracy of the electricity load prediction model of the building individual.

7. The parameterized characterization method as described in claim 6, characterized in that, The extracted orderliness indicators for individual buildings include: The electricity load curves of the individual buildings are converted from the time domain to the frequency domain to obtain the electricity load spectrum diagram; Extract the amplitude of the main frequency components in the predetermined period of the power load spectrum diagram; Calculate the ratio of the sum of the amplitudes of the main frequency components in the predetermined period to the total spectral amplitude of the power load spectrum, and use the ratio as the orderliness index.

8. The parameterized characterization method as described in claim 6 or 7, characterized in that, The method further includes: Based on the first mapping relationship and the second mapping relationship, building groups and individual buildings are classified according to the prediction difficulty; Based on the difficulty level of prediction, corresponding electricity consumption data filtering strategies are adopted.

9. The parameterized characterization method as described in claim 8, characterized in that, The classification of building groups and individual buildings according to prediction difficulty includes: Building groups and individual buildings are classified into three categories according to the difficulty of prediction: easy to predict, medium difficulty to predict, and difficult to predict. The data filtering strategy based on the prediction difficulty classification includes: For buildings with easily predictable electricity loads, a unified electricity load prediction model is used to predict electricity loads in batches. For buildings requiring moderate-difficulty forecasting, the structure of the electricity load forecasting model will be further optimized. For buildings that are difficult to predict, increase the sampling frequency, improve data quality, or build personalized electricity load prediction models.

10. The parameterized characterization method as described in claim 8, characterized in that, The method further includes: Assign prediction difficulty level labels to all buildings for easy subsequent use.

Images