Method for public building energy consumption quota based on multi-level classification and dynamic data correction

Abstract

The application discloses a kind of based on multi-level classification and dynamic data correction public building energy consumption quota method, comprising: obtaining energy consumption parameter, constructs multi-feature dataset and pre-processes;Based on functionality and energy consumption attribute, first classification is carried out;Based on K-means model, secondary classification is carried out;Based on quota level and sorting coupling method and simulation quota method, energy consumption quota is carried out respectively and weighted average is obtained to obtain comprehensive energy consumption quota;Based on SHAP value model, energy consumption feature importance analysis is carried out;Based on the annual data of important features, linear regression is carried out, and important feature prediction value and change rate are calculated to correct the comprehensive energy consumption quota of each type of building in the future;Based on the change rate of the important features of single building, energy use rationality and energy-saving enthusiasm are judged, and the comprehensive energy consumption quota is corrected.The application is based on dynamic data to correct two kinds of energy consumption quota of multi-level classification building, which improves the influence of the previous energy consumption quota method on building energy consumption, which is ignored by different behaviors, properties and characteristics between buildings.

Description

Technical Field

[0001] This invention belongs to the technical field of computer vision, specifically relating to a method for determining energy consumption quotas for public buildings based on multi-level classification and dynamic data correction. Background Technology

[0002] Improving energy efficiency during the operation of public buildings is necessary. Energy consumption quotas refer to the upper limit of the total energy consumed by a building to maintain its equipment, facilities, and sanitary conditions within a quota period, and serve as a benchmark for measuring the level of energy consumption in buildings.

[0003] A scientifically sound energy consumption quota method forms the basis for assessing and evaluating the energy management level of public institutions, and is also a crucial tool for achieving the "dual control" targets for total energy consumption and intensity in public institutions. However, differences in operating modes and building characteristics among different buildings limit the rationality of individual building energy consumption benchmarks, necessitating adjustments to these benchmarks based on the specific characteristics of different public buildings. This requires the establishment of a vast national database and the statistical evaluation of historical operating data from a large number of different types of public buildings. Although numerous energy consumption quota methods exist, current methods still suffer from numerous inconsistencies due to the influence of geographical location, climate, and energy-consuming personnel on energy consumption among similar buildings in China. Therefore, it is worth considering how to most rationally divide public building groups and eliminate the impact of different factors on building energy consumption. Summary of the Invention

[0004] The main objective of this invention is to overcome the shortcomings and deficiencies of existing technologies and provide a method for energy consumption quotas of public buildings based on multi-level classification and dynamic data correction. The method combines multi-level public building type classification and quota level with a ranking coupling method and a simulation quota method to obtain a comprehensive energy consumption quota value. The comprehensive energy consumption quota value is then reasonably corrected based on dynamic data, providing a more reasonable energy consumption quota method for government building energy consumption quota setters.

[0005] To achieve the above objectives, the present invention adopts the following technical solution:

[0006] One aspect of the present invention provides a method for energy consumption quotas for public buildings based on multi-level classification and dynamic data correction, comprising the following steps:

[0007] Obtain monthly energy consumption parameters for different types of public buildings over several years, including date parameters, load parameters, and energy-saving measure parameters;

[0008] Based on the obtained energy consumption parameters, a multi-feature dataset is constructed and data verification and cleaning, data feature filtering, and data normalization are performed.

[0009] Based on the functional and energy-use attributes of public buildings, a first-level logical classification and labeling of a multi-feature dataset is performed;

[0010] The results of the first-level logical and classification labels are used to perform second-level classification based on the K-means model, and the results of the second-level classification are logically interpreted and labeled according to feature information;

[0011] Energy consumption quotas were set for public buildings of the same secondary type based on the quota level and ranking coupling method and the simulation quota method, and the quota values ​​obtained by the two methods were weighted and averaged to obtain the comprehensive energy consumption quota.

[0012] Based on the SHAP value model, the importance of annual energy consumption characteristics of different types of public buildings is analyzed, and the important characteristics and their Shapley value weights are obtained.

[0013] Linear regression was performed based on the annual average data of the key features of each type of public building, and the average value and rate of change of the key features of each type of public building in the coming years were predicted based on the regression curve.

[0014] The comprehensive energy consumption quota for various types of buildings in the next few years is obtained by weighting and correcting the rate of change of important features and the Shapley value of important features.

[0015] The rate of change over the past few years is obtained by linear regression of monthly data on key characteristics of a single building. This is used to determine the rationality of energy use and the initiative for energy conservation in a single building, and to revise the comprehensive energy consumption quota of a single building.

[0016] As a preferred technical solution, the different types of public buildings include educational buildings, administrative buildings, and medical buildings; the date parameters include date, year, month, and quarter; the load parameters include per capita water consumption, building area, number of energy users, authorized staffing, number of gasoline vehicles, number of diesel vehicles, number of new energy vehicles, electricity consumption, coal consumption, natural gas consumption, oil consumption, other oil consumption, diesel consumption, liquefied petroleum gas consumption, other energy consumption, solar collector area, solar system installed capacity, and geothermal system installed capacity; the energy-saving measures parameters include whether energy-saving renovations have been carried out, the duration since the energy-saving renovations, and the energy-saving renovation projects carried out.

[0017] As a preferred technical solution, the data verification and cleaning specifically involves: using the quartile method to verify and clean the acquired energy consumption parameters, specifically: taking Q3 + 1.5IQR of the distribution interval as the upper threshold and Q1 - 1.5IQR as the lower threshold; where Q3 is the third quartile of the distribution interval, that is, the number at the 75th percentile after arranging the data sequence from smallest to largest; Q1 is the first quartile of the distribution interval, that is, the number at the 25th percentile after arranging the data sequence from smallest to largest; IQR is the interquartile range, IQR = Q3 - Q1, and values ​​outside Q1 to Q3 are discarded as outliers;

[0018] The data feature screening specifically involves: using the random forest method to analyze the correlation between per capita energy consumption and energy consumption per unit area of ​​public buildings and each influencing factor in the influencing factor sequence S, screening out influencing factors with a correlation higher than the preset value, and deleting influencing factors with a correlation lower than the preset value; wherein the influencing factor sequence S = [year, month, quarter, per capita water consumption PWC, building area BA, number of energy users NEU, number of authorized personnel H, number of gasoline vehicles NGV, number of diesel vehicles NDV, petroleum consumption GC, other oil consumption OC, diesel consumption DC, liquefied petroleum consumption CLP, other energy consumption OEC, solar collector area SCA, solar system installed capacity SSIC, geothermal system installed capacity GIC, energy-saving renovation ESR];

[0019] The data normalization process is specifically as follows:

[0020]

[0021] Where X is the data to be processed, X max X is the maximum value in the dataset. min X is the minimum value in the dataset. n This is the normalized data.

[0022] As a preferred technical solution, the method involves performing secondary classification on the results of the primary logical and classification labels based on the K-means model, and then logically interpreting and labeling the results of the secondary classification based on feature information. Specifically:

[0023] Clustering of multi-feature datasets of different types of public buildings is performed based on the K-means model, and the K value of samples of different types of public buildings is determined according to the "elbow method". The "elbow method" specifically involves: calculating the sum of squared errors within clusters, finding the optimal number of classifications using image inflection points, and then performing cluster analysis based on the K-means model and input data. The K value is the number of subcategories after the secondary classification of different types of public buildings. The input of the K-means model is energy consumption data directly related to annual energy consumption, including annual energy consumption, annual per capita energy consumption, annual energy consumption per unit area, annual electricity consumption, annual oil consumption, and annual natural gas consumption.

[0024] The results of the secondary classification of different types of public buildings are logically explained as follows: the specific locations of all types of public buildings are marked on the map and analyzed; the cluster points of unknown markers for the same type of secondary public building are described, and the labels of the secondary public buildings are obtained accordingly.

[0025] As a preferred technical solution, the comprehensive energy consumption quota calculation method is as follows:

[0026]

[0027] EUI C EUI is the comprehensive energy consumption quota value. o EUI is the original comprehensive energy consumption quota calculated using the quota level and ranking coupling method based on the original data. p The simulated comprehensive energy consumption quota is calculated using the simulated quota method;

[0028] The specific method of coupling quota levels and sorting is as follows:

[0029] To determine whether a certain type of sample conforms to a normal distribution, if it does, the quota level method is used to calculate the comprehensive energy consumption quota for the original data, and 20%, 50%, and 80% of the quota level are selected as the constraint value, benchmark value, and guiding value of the energy consumption quota, respectively; if it does not conform, the data in the sample are sorted, and the proportions of the sorted data are selected as the constraint value, benchmark value, and guiding value of the energy consumption quota, respectively.

[0030] The simulated quota method specifically refers to:

[0031] The multi-feature dataset is divided into training and test sets according to a certain ratio;

[0032] During the training phase, each model is trained based on a prediction model, and the hyperparameters of each model are adjusted using optimization algorithms. The prediction models include stepwise linear regression, support vector machine, random forest, gradient boosting, and backpropagation neural network. The optimization algorithms include nonlinear weighted particle swarm optimization, whale algorithm, and sparrow algorithm.

[0033] During the testing phase, the annual per capita energy consumption and annual energy consumption per unit area for each type of public building are predicted. Once the prediction results are evaluated by preset indicators, the energy consumption prediction model is considered to have been trained. Through indicator comparison, a prediction model and optimization algorithm suitable for the current type of public building are selected, and the simulated comprehensive energy consumption quota for each type of public building is predicted.

[0034] As a preferred technical solution, the eigenvector correlation analysis of the annual energy consumption of various types of public buildings based on the SHAP value model is specifically as follows:

[0035] Based on the SHAP value model, the contribution of energy consumption parameters of different types of public buildings is calculated to obtain the contribution order of all features. Specifically, the marginal contribution of a feature when it is added to the model is calculated, and the different marginal contributions of the feature in all feature sequences are considered. The average value is taken to obtain the Shapley value weight of the feature.

[0036] Important features are selected based on the interaction between features and the order of contribution of features that are sensitive to the results.

[0037] As a preferred technical solution, the comprehensive energy consumption quota for various types of buildings in the next few years is obtained by weighted correction based on the rate of change of important features and the Shapley value of important features, specifically as follows:

[0038] PCEC next =PCEC now *(a1*ω1+a2*ω2+a3*ω3);

[0039] Among them, PCEC next For a future year, the PCEC set a per capita energy consumption quota for this type of building. now ω1 represents the current per capita energy consumption quota for this type of building, a1 represents the percentage decrease of the first eigenvector after several years, ω1 represents the proportion of the Shapley value of the first eigenvector in the total Shapley value of the three eigenvectors, and so on.

[0040] As a preferred technical solution, the rate of change over the past few years obtained by linear regression of monthly data based on the key characteristics of a single building is used to determine the rationality of energy use and energy-saving initiative of the single building, and to revise the comprehensive energy consumption quota of the single building. Specifically:

[0041] If any abnormal data points in the past few years' change curve of a building's important features exceed the normal fluctuation range by 20%, or if the characteristic value shows an increasing trend with an increase of more than 20%, it is judged as unreasonable energy use, and the comprehensive energy consumption quota is reduced by 5%. If the past few years' change trend of a building's important features shows a decreasing trend with a decrease of more than 20%, it is judged as actively responding to the energy conservation and emission reduction policy, and the comprehensive energy consumption quota is increased by 5%.

[0042] Another aspect of the present invention provides a system for energy consumption quotas of public buildings based on multi-level classification and dynamic data correction, which is applied to the above-mentioned method for energy consumption quotas of public buildings based on multi-level classification and dynamic data correction. The system includes a data acquisition and processing module, a data preprocessing module, a building type primary classification module, a building type secondary classification module, an energy consumption quota module, a feature importance analysis module, and an energy consumption quota value correction module.

[0043] The data acquisition and processing module is used to obtain monthly and annual energy consumption parameters for different types of shared buildings, including date parameters, load parameters, and energy-saving measure parameters; and to construct a multi-feature dataset based on the obtained energy consumption parameters.

[0044] The data preprocessing module is used to perform data verification and cleaning, data feature filtering, and data normalization on multi-feature datasets.

[0045] The building type primary classification module is used to classify and label the sample set according to the functionality and energy consumption attributes of public buildings.

[0046] The building type secondary classification module is used to perform secondary classification based on the results of the primary logic and classification labels according to the K-means model, and to provide a logical interpretation and label of the secondary classification results based on feature information;

[0047] The energy consumption quota module is used to calculate energy consumption quotas for public buildings of the same secondary type according to the quota level and ranking coupling method and the simulation quota method, and to calculate the comprehensive energy consumption quota by weighted averaging of the quota values ​​obtained by the two methods.

[0048] The feature importance analysis module is used to perform annual energy consumption feature importance analysis on different types of public buildings based on the SHAP value model, according to energy consumption parameters, energy consumption prediction model and energy consumption prediction results, to obtain important features and the Shapley value weight of important features.

[0049] The energy consumption quota correction module is used to perform linear regression based on the annual average data of the important characteristics of each type of public building, and predict the average value and rate of change of the important characteristics of each type of public building in the next few years based on the regression curve.

[0050] The comprehensive energy consumption quota for various types of buildings in the next few years is obtained by weighting and correcting the rate of change of important features and the Shapley value of important features.

[0051] The rate of change over the past few years is obtained by linear regression of monthly data on key characteristics of a single building. This is used to determine the rationality of energy use and the initiative for energy conservation in a single building, and to revise the comprehensive energy consumption quota of a single building.

[0052] In another aspect, the present invention also provides a storage medium storing a program that, when executed by a processor, implements the above-described method for determining energy consumption quotas for public buildings based on multi-level classification and dynamic data correction.

[0053] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0054] (1) In the field of energy consumption quota setting for public buildings, this invention realizes the energy consumption quota per unit area and per capita energy consumption for different types of public buildings, takes into account more energy consumption influencing factors and makes multi-level reasonable classification, and the quota value is more reasonable; it can predict the energy consumption quota value for future years based on dynamic data, provide direction for different types of public buildings on whether they need energy-saving renovation, and provide support for building energy management and energy conservation; the data verification and missing data filling of this invention can be completed automatically, and the effective dataset can be constructed. The processing speed is fast and the operation is convenient. Attached Figure Description

[0055] Figure 1 This is a flowchart illustrating a method for correcting public building energy consumption quotas based on multi-level classification and dynamic data in an embodiment of the present invention.

[0056] Figure 2 This is a flowchart illustrating the primary and secondary classification methods for public buildings in this embodiment of the invention.

[0057] Figure 3 This is a flowchart illustrating the energy consumption quotas for various building categories in an embodiment of the present invention.

[0058] Figure 4 This is a schematic diagram illustrating a process for correcting comprehensive energy consumption quotas based on dynamic data, according to an embodiment of the present invention.

[0059] Figure 5 This is a schematic diagram of the structure of a system for correcting energy consumption quotas for public buildings based on dynamic data, according to an embodiment of the present invention.

[0060] Figure 6 This is a schematic diagram of the structure of the storage medium according to an embodiment of the present invention. Detailed Implementation

[0061] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of the present application, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present application without creative effort are within the scope of protection of the present application.

[0062] It should be noted that the following detailed descriptions are exemplary and intended to provide further explanation of this application. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains.

[0063] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of exemplary embodiments according to the invention. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0064] Where there is no conflict, the embodiments and features in the embodiments of the present invention can be combined with each other.

[0065] like Figure 1 As shown, Embodiment 1 of the present invention discloses a method for correcting the energy consumption quota of public buildings based on multi-level classification and dynamic data, including the following steps:

[0066] S1. Obtain annual and monthly energy consumption parameters for different types of public buildings with energy consumption quotas over several years, including date parameters, load parameters, and energy-saving measure parameters.

[0067] As one or more implementation methods, the public buildings to be assessed include at least educational buildings, administrative buildings, stadiums, and medical buildings; the date parameters include at least date, year, month, and quarter; the load parameters include at least per capita water consumption (PWC), building area (BA), number of energy users (NEU), authorized staff (H), number of gasoline vehicles (NGV), number of diesel vehicles (NDV), number of new energy vehicles (NNV), electricity consumption (EC), coal consumption (CC), natural gas consumption (NGC), petroleum consumption (GC), other oil consumption (OC), diesel consumption (DC), liquefied petroleum gas consumption (CLP), other energy consumption (OEC), solar collector area (SCA), solar system installed capacity (SSIC), and geothermal system installed capacity (GIC). The energy-saving measures parameters include at least whether energy-saving renovations have been carried out (what type of renovation), how long since the renovation, and the specific energy-saving renovation projects carried out. Electricity and water consumption information is obtained through electricity and water meters, while other energy consumption information and energy-saving renovation information are obtained through manual reporting.

[0068] S2. Construct a multi-feature dataset based on the obtained energy consumption parameters and perform data verification and cleaning, data feature filtering, and data normalization.

[0069] Furthermore, in this embodiment, the construction of a multi-feature dataset based on the acquired energy consumption parameters specifically involves:

[0070] S21. Since the acquired parameters may contain abnormal or missing data, in this embodiment, abnormal data needs to be checked for load parameters to avoid outliers, and missing data needs to be filled using a preliminary calculation method; for date data, more dimensional features need to be extracted; for energy-saving measure data, it needs to be judged and corrected in conjunction with the energy consumption changes over the years. Specifically:

[0071] S22. Data Verification and Cleaning: The obtained energy load parameters are verified and cleaned using the quartile method. Specifically, Q3 + 1.5IQR is taken as the upper threshold and Q1 - 1.5IQR as the lower threshold. Here, Q3 is the third quartile of the distribution interval, which is the number at the 75th percentile after arranging the data sequence from smallest to largest; Q1 is the first quartile of the distribution interval, which is the number at the 25th percentile after arranging the data sequence from smallest to largest; IQR is the interquartile range, IQR = Q3 - Q1, and values ​​outside Q1 to Q3 are discarded as outliers.

[0072] Extract the year, month, and quarter for subsequent data analysis and processing to provide more accurate and detailed information on energy consumption changes and trends.

[0073] Finally, the validated data is formatted to ensure its standardization and consistency. Boolean data uses "0" to represent "False" and "1" to represent "True". The final sequence S contains the following elements:

[0074] S=[year,month,quarter,PWC,BA,NEU,H,NGV,NDV,GC,OC,DC,CLP,OEC,SCA,SSIC,GIC,ESR].

[0075] Specifically, in cases of missing historical data, alternative data is generated based on the same period's historical data of similar energy-consuming units or the previous period's historical data of the target energy-consuming unit to fill in the missing reporting values; in cases of missing data to be verified, the missing reporting values ​​for the current month are filled in after verification with the target energy-consuming unit.

[0076] S23. Data Feature Filtering:

[0077] Because the correlation between per capita energy consumption and energy consumption per unit area of ​​public buildings and the elements in sequence S is different, to obtain more accurate prediction results, per capita energy consumption and energy consumption per unit area are predicted separately. Therefore, for the two energy consumption intensities, data with higher correlation need to be selected as input to the model. This can reduce the difficulty of model training and improve the accuracy of the model. Specifically, the following steps are included:

[0078] The random forest model was used to analyze the correlation between per capita energy consumption and energy consumption per unit area of ​​public buildings and various influencing factors in sequence S. Influencing factors with a correlation higher than the preset value were screened out, and influencing factors with a correlation lower than the preset value were deleted.

[0079] S24. Data normalization processing:

[0080] To eliminate the influence of dimensions and orders of magnitude among different features, and to make the impact of each feature on the results more balanced, preventing certain features from dominating due to excessively large or small value ranges, thereby reducing the accuracy of the model, normalization is performed on all data. This makes the contribution of each feature more equitable and improves the accuracy of the model, as shown in the following formula:

[0081]

[0082] Where X is the data to be processed, X max X is the maximum value in the dataset. min X is the minimum value in the dataset. n This is the normalized data.

[0083] S3. Based on the functionality and energy consumption attributes of public buildings, classify and label each type of public building into primary logical categories.

[0084] like Figure 2 As shown, as one or more implementation methods, this embodiment classifies and labels the sample set based on the functionality and energy consumption attributes of public buildings. For example, educational buildings are divided into kindergartens, primary schools, junior high schools, senior high schools and adult education institutions; medical buildings are divided into primary hospitals, secondary hospitals and tertiary hospitals; and administrative buildings are divided into administrative buildings using split air conditioners, administrative buildings using centralized air conditioners and other administrative buildings.

[0085] S4. Based on the K-means model, perform secondary classification on the results of the first-level logical and classification labels, and logically interpret and label the results of the secondary classification according to the feature information.

[0086] As one or more implementation methods, this embodiment uses the K-means model to cluster sample sets of different types of public buildings. The model inputs energy consumption data directly related to annual energy consumption, such as annual energy consumption, annual per capita energy consumption, annual energy consumption per unit area, annual electricity consumption, annual oil consumption, and annual natural gas consumption. The K value of different types of public building samples is determined according to the "elbow method", that is, the sum of squared errors within the cluster (SSE) is calculated. The optimal number of classifications is found by using image inflection points. Then, cluster analysis is performed based on the model and the input data. The K value is the number of subcategories after the secondary classification of different types of public buildings.

[0087] Furthermore, the results of the secondary classification of public building types are logically interpreted. The specific locations of all types of public buildings are marked and analyzed on the map. The cluster points of unknown markers for the same secondary type of public building are described, and the labels of secondary type public buildings are obtained, such as district boundary primary school, district central primary school, etc.

[0088] S5. Energy consumption quotas are set for public buildings of the same secondary type based on the quota level and sorting coupling method.

[0089] Furthermore, the energy consumption quotas for the original data of public buildings belonging to the same secondary type are as follows:

[0090] like Figure 3 As shown, the quota level and ranking coupling method first determines whether a certain type of sample conforms to a normal distribution or is similar to it. If it does, the quota level method is used to set energy consumption quotas for the sample, selecting 20%, 50%, and 80% of the quota level as constraint values, benchmark values, and guiding values ​​for the energy consumption quota, respectively. If it does not conform, the data in the sample is ranked, and the proportions in the ranking are selected as 20%, 50%, and 80% as constraint values, benchmark values, and guiding values ​​for the energy consumption quota, respectively.

[0091] S6. Energy consumption quotas are set for public buildings of the same secondary type based on the simulation quota method.

[0092] The simulated quota method divides the multi-feature dataset into a training set and a test set. In one or more preferred embodiments, 70% of the multi-feature dataset is used to construct the training set, and 30% is used to construct the test set.

[0093] During the training phase, each model is trained based on a prediction model, and the hyperparameters of each model are adjusted by an optimization algorithm. Preferably, in this embodiment, the prediction model includes stepwise linear regression, support vector machine, random forest, gradient boosting method and BP neural network, and the optimization algorithm includes nonlinear weight decreasing particle swarm optimization algorithm, whale algorithm and sparrow algorithm.

[0094] During the testing phase, annual per capita energy consumption and annual energy consumption per unit area were predicted for each type of building. Once the prediction results passed the evaluation of preset indicators, the energy consumption prediction model was considered to have been successfully trained. Through indicator comparison, the prediction model and optimization algorithm most suitable for the current type of public building were selected.

[0095] The energy consumption indicators of various types of public buildings are predicted based on the prediction model and optimization algorithm most suitable for the current type of public building.

[0096] S7. Calculation steps of the repetitive quota level of predicted data and the sorting coupling method based on the simulation quota method;

[0097] Furthermore, the weighted average of the quota values ​​obtained from the two methods yields the comprehensive energy consumption quota, which is calculated using the following formula:

[0098]

[0099] EUI C EUI is a comprehensive energy consumption quota. o EUI is the energy consumption quota calculated based on the raw data. p This is the energy consumption quota value calculated based on simulation data.

[0100] S8. Based on the SHAP value model, perform eigenvector correlation analysis on the annual energy consumption of various types of public buildings, specifically as follows:

[0101] The contribution of energy consumption parameters of different types of public buildings is calculated based on the SHAP value method, and the contribution order of all features is obtained. The SHAP value model is an interpretable contribution analysis method based on model training. It calculates the contribution of each feature to the prediction value. Specifically, it calculates the marginal contribution of a feature when it is added to the model, considers the different marginal contributions of the feature in all feature sequences, takes the average value, and obtains the Shapley value weight of the feature.

[0102] Important features are selected based on the interaction between features and the order of contribution of features that are sensitive to the results.

[0103] S9. Analyze the annual and monthly trends of key characteristics, and analyze the annual trends of annual energy consumption quota per unit area and annual per capita energy consumption quota. Revise the comprehensive energy consumption quota and predict the energy consumption quota for future years. Specifically:

[0104] like Figure 4 As shown, a linear regression is performed on the annual average data of key characteristics for each type of public building. Based on the regression curve, the average value and rate of change of key characteristics for each type of public building over the next few years are predicted. Then, the comprehensive energy consumption of different types of public buildings is corrected based on the rate of change of key characteristics and the weight of Shapley values. The correction formula is as follows:

[0105] PCEC next =PCEC now *(a1*ω1+a2*ω2+a3*ω3);

[0106] Among them, PCEC next For a future year, the PCEC set a per capita energy consumption quota for this type of public building. nowThis represents the current per capita energy consumption quota for this type of public building. a1 represents the percentage decrease of the first eigenvector after several years. ω1 represents the proportion of the Shapley value of the first eigenvector in the total Shapley value of the three eigenvectors, and so on.

[0107] Furthermore, based on monthly data of key characteristics of individual buildings, linear regression is performed to obtain the rate of change over the past few years. This is used to assess the rationality of energy use and the initiative in energy conservation of individual buildings, and to revise the comprehensive energy consumption quota of individual buildings, as follows:

[0108] The method for judging the rationality of energy use and the initiative in energy conservation of a single building is as follows: if there are abnormal data points in the past few years' change curve of a building's important characteristics that exceed the normal fluctuation range by 20%, then the energy use is unreasonable and a penalty of 5% reduction in the comprehensive energy consumption quota is imposed; if the change trend of a building's important characteristics over the past few years is a decreasing trend and the decrease is greater than 20%, it indicates that the building is actively responding to the energy conservation and emission reduction policy and a reward of 5% increase in the comprehensive energy consumption quota is imposed; conversely, if the change trend of the characteristic value over the past few years is an increasing trend and the increase is greater than 20%, a penalty of 5% reduction in the comprehensive energy consumption quota is imposed.

[0109] Example 2

[0110] like Figure 5 As shown, in another embodiment of this application, a system for energy consumption quotas of public buildings based on multi-level classification and dynamic data correction is provided. The system includes a data acquisition and processing module, a data preprocessing module, a building type primary classification module, a building type secondary classification module, an energy consumption quota module, a feature importance analysis module, and an energy consumption quota value correction module.

[0111] The data acquisition and processing module is used to obtain monthly and annual energy consumption parameters for different types of public buildings, including date parameters, load parameters, and energy-saving measure parameters; and to construct a multi-feature dataset based on the obtained energy consumption parameters.

[0112] The data preprocessing module is used to perform data verification and cleaning, data feature filtering, and data normalization on multi-feature datasets.

[0113] The building type primary classification module is used to classify and label the sample set according to the functionality and energy consumption attributes of public buildings.

[0114] The building type secondary classification module is used to perform secondary classification based on the results of the primary logic and classification labels according to the K-means model, and to provide a logical interpretation and label of the secondary classification results based on feature information;

[0115] The energy consumption quota module is used to calculate energy consumption quotas for public buildings of the same secondary type according to the quota level and ranking coupling method and the simulation quota method, and to calculate the comprehensive energy consumption quota by weighted averaging of the quota values ​​obtained by the two methods.

[0116] The feature importance analysis module is used to perform annual energy consumption feature importance analysis on different types of public buildings based on the SHAP value model, according to energy consumption parameters, energy consumption prediction model and energy consumption prediction results, to obtain important features and the Shapley value weight of important features.

[0117] The energy consumption quota correction module is used to perform linear regression based on the annual average data of the important characteristics of each type of public building, and predict the average value and rate of change of the important characteristics of each type of public building in the next few years based on the regression curve.

[0118] The comprehensive energy consumption quota for various types of public buildings in the coming years is derived by weighting and correcting the rate of change of important features and the Shapley value of important features.

[0119] The rate of change over the past few years is obtained by linear regression of monthly data on key characteristics of a single building. This is used to determine the rationality of energy use and the initiative for energy conservation in a single building, and to revise the comprehensive energy consumption quota of a single building.

[0120] It should be noted that the system provided in the above embodiments is only an example of the division of the above functional modules. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure can be divided into different functional modules to complete all or part of the functions described above. This system is a method for energy consumption quota of public buildings based on multi-level classification and dynamic data correction applied to the above embodiments.

[0121] Example 3

[0122] like Figure 6 As shown, in another embodiment of this application, a computer-readable storage medium and an electronic device are also provided. The computer-readable storage medium stores a program, which, when executed by the processor of the electronic device, implements a method for correcting public building energy consumption quotas based on multi-level classification and dynamic data, as described in the above embodiment. Specifically, the method is as follows:

[0123] Obtain monthly energy consumption parameters for different types of public buildings over several years, including date parameters, load parameters, and energy-saving measure parameters;

[0124] Based on the obtained energy consumption parameters, a multi-feature dataset is constructed and data verification and cleaning, data feature filtering, and data normalization are performed.

[0125] Based on the functional and energy-use attributes of public buildings, a first-level logical classification and labeling of a multi-feature dataset is performed;

[0126] The results of the first-level logical and classification labels are used to perform second-level classification based on the K-means model, and the results of the second-level classification are logically interpreted and labeled according to feature information;

[0127] Energy consumption quotas were set for public buildings of the same secondary type based on the quota level and ranking coupling method and the simulation quota method, and the quota values ​​obtained by the two methods were weighted and averaged to obtain the comprehensive energy consumption quota.

[0128] Based on the SHAP value model, the importance of annual energy consumption characteristics of different types of public buildings is analyzed, and the important characteristics and their Shapley value weights are obtained.

[0129] Linear regression was performed based on the annual average data of the key features of each type of public building, and the average value and rate of change of the key features of each type of public building in the coming years were predicted based on the regression curve.

[0130] The comprehensive energy consumption quota for various types of buildings in the next few years is obtained by weighting and correcting the rate of change of important features and the Shapley value of important features.

[0131] The rate of change over the past few years is obtained by linear regression of monthly data on key characteristics of a single building. This is used to determine the rationality of energy use and the initiative for energy conservation in a single building, and to revise the comprehensive energy consumption quota of a single building.

[0132] It should be understood that various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0133] The above embodiments are preferred embodiments of the present invention, but the embodiments of the present invention are not limited to the above embodiments. Any changes, modifications, substitutions, combinations, or simplifications made without departing from the spirit and principle of the present invention shall be considered equivalent substitutions and shall be included within the protection scope of the present invention.

Claims

1. A method for public building energy consumption quota based on multi-level classification and dynamic data correction, characterized in that, The method comprises the following steps: obtaining monthly energy consumption parameters of different types of public buildings in several years, including date parameters, load parameters and energy-saving measure parameters; constructing a multi-feature data set according to the obtained energy consumption parameters and performing data verification and cleaning, data feature screening and data normalization processing; performing first-level logical classification and labeling on the multi-feature data set based on the functionality and energy consumption properties of public buildings; performing second-level classification based on the K-means model on the results of the first-level logical classification and labeling, and logically explaining and labeling the results of the second-level classification based on feature information; performing energy consumption quota on public buildings of the same second-level type based on the quota level and sorting coupling method and the simulation quota method, and weighting the quota values obtained by the two methods to obtain a comprehensive energy consumption quota; performing annual energy consumption feature importance analysis on different types of public buildings based on the SHAP value model, and obtaining important features and Shapley value weights of important features, specifically: performing feature contribution degree calculation on energy consumption parameters of different types of public buildings based on the SHAP value model, and obtaining the contribution degree sequence results of all features, specifically: calculating the marginal contribution of a feature when it is added to the model, and considering the different marginal contributions of the feature in all feature sequences, and taking the average to obtain the Shapley value weight of the feature; screening important features according to the contribution degree sequence results of the features and the features sensitive to the results; performing linear regression based on the annual data average value of the important features of each type of public building, and predicting the average value and change rate of the important features of each type of public building in the future several years according to the regression curve; obtaining the comprehensive energy consumption quota of each type of building in the future several years based on the change rate of the important features and the Shapley value weight of the important features, specifically: PCEC next = PCEC now *( a 1* ω 1+ a 2* ω 2+ a 3* ω 3)； wherein, PCEC next the future per capita energy consumption quota value of such buildings in a certain year, PCEC now represent the current per capita energy consumption quota value of such buildings, a 1、 a 2、 a 3 respectively represent the proportion of the first, second and third characteristic vectors after a number of years, ω 1、 ω 2、 ω 3 respectively represent the proportion of the Shapley values of the first, second and third characteristic vectors in the total Shapley values of the three characteristic vectors; performing linear regression based on the monthly data of the important features of a single building to obtain the change rate in the past several years, judging the energy consumption rationality and energy-saving enthusiasm of the single building, and correcting the comprehensive energy consumption quota of the single building.

2. The method for public building energy consumption quota based on multi-level classification and dynamic data correction according to claim 1, characterized in that, The different types of public buildings include educational buildings, administrative buildings and medical buildings; the date parameters include date, year, month and season; the load parameters include water consumption per capita, building area, energy consumption number, designed number, gasoline vehicle number, diesel vehicle number, new energy vehicle number, power consumption, coal consumption, natural gas consumption, oil consumption, other oil consumption, diesel consumption, liquefied petroleum consumption, other energy consumption, solar collector area, solar system installed capacity and geothermal system installed capacity; the energy-saving measure parameters include whether energy-saving reconstruction is performed, the time length from energy-saving reconstruction to the present and the energy-saving reconstruction project performed.

3. The method for public building energy consumption quota based on multi-level classification and dynamic data correction according to claim 1, characterized in that, The data verification and cleaning specifically comprises: using the quartile method to verify and clean the obtained energy consumption parameters, specifically: taking Q3+1.5IQR of the distribution interval as the upper threshold, and taking Q1-1.5IQR as the lower threshold; wherein Q3 is the third quartile of the distribution interval, that is, the number at the 75% position after arranging the data sequence from small to large; Q1 is the first quartile of the distribution interval, that is, the number at the 25% position after arranging the data sequence from small to large; IQR is the interquartile range, IQR=Q3-Q1, and the values outside Q1-Q3 are discarded as outliers; The data feature screening specifically comprises: using the random forest method to analyze the correlation between the per capita energy consumption and unit area energy consumption of the public building and each influencing factor in the influencing factor sequence S, and screening out the influencing factors with a correlation higher than a preset value, and deleting the influencing factors with a correlation lower than the preset value; wherein the influencing factor sequence S=[year, month, quarter, PWC, BA, NEU, H, NGV, NDV, GC, OC, DC, CLP, OEC, SCA, SSIC, GIC, ESR]; The normalization processing of the data is specifically as follows: ； wherein, X is the data to be processed, X max is the maximum value in the data, X min is the minimum value in the data, X n is the normalized data.

4. The method for public building energy consumption quota based on multi-level classification and dynamic data correction according to claim 1, characterized in that, The K-means model is used to perform secondary classification on the results of the first-level logic and classification labels, and the results of the secondary classification are logically explained and labeled according to feature information, specifically: The K-means model is used to cluster the multi-feature data sets of different types of public buildings, and the K value of the different types of public building samples is determined according to the "elbow method"; the "elbow method" specifically comprises: calculating the within-cluster sum of squares of errors, finding the best classification number by using the image inflection point, and then performing clustering analysis based on the K-means model and the input data, and the K value is the number of subcategories after the secondary classification of different types of public buildings; the K-means model inputs energy consumption data directly related to annual energy consumption, including annual energy consumption, annual per capita energy consumption, annual unit area energy consumption, annual power consumption, annual oil consumption, and annual natural gas consumption; The results of the secondary classification of different types of public buildings are logically explained, specifically: marking the specific positions of all types of public buildings on the map and analyzing, describing the unknown label aggregation points based on the same secondary type of public building, and obtaining the label of the secondary type of public building.

5. The method for public building energy consumption quota based on multi-level classification and dynamic data correction according to claim 1, characterized in that, The comprehensive energy consumption quota calculation method is as follows: ； wherein EUI C is a comprehensive energy consumption quota value, EUI O is a raw comprehensive energy consumption quota calculated by using the quota level and sorting coupling method based on raw data, EUI p is a simulation comprehensive energy consumption quota calculated by using the simulation quota method. Wherein the quota level and sorting coupling method specifically comprises: Determine whether a type of sample conforms to normal distribution, if so, adopt the quota level method to calculate the comprehensive energy consumption quota of the original data, and select 20%, 50% and 80% of the quota level as the constraint value, the benchmark value and the guide value of the energy consumption quota respectively; if not, sort the data in the sample, and select 20%, 50% and 80% of the proportion in the sorting as the constraint value, the benchmark value and the guide value of the energy consumption quota respectively; The simulation quota method specifically comprises the following steps: Divide the multi-feature data set into a training set and a test set according to a proportion; In the training stage, each model is trained based on a prediction model, and the hyperparameters of each model are adjusted by an optimization algorithm; wherein the prediction model includes stepwise linear regression, support vector machine, random forest, gradient boosting method and BP neural network, and the optimization algorithm includes nonlinear weight decreasing particle swarm optimization algorithm, whale algorithm and sparrow algorithm; In the test stage, the annual energy consumption per capita and the annual energy consumption per unit area of each type of public building are predicted respectively, and when the prediction result passes the preset index evaluation, it is considered that the energy consumption prediction model is trained. Through index comparison, the prediction model and optimization algorithm suitable for the current type of public building are selected to predict the simulation comprehensive energy consumption quota of each type of public building.

6. The method for public building energy consumption quota based on multi-level classification and dynamic data correction according to claim 1, characterized in that, The change rate of the past several years obtained by linear regression based on the monthly data of the important features of a single building is used to determine the energy use rationality and energy saving enthusiasm of the single building, and the comprehensive energy consumption quota of the single building is corrected, specifically as follows: If there are abnormal data points in the past several years of the change curve of the important features of a building, which are more than 20% of the normal fluctuation range, or the feature value change trend shows an increasing trend and the rising amplitude is greater than 20%, it is judged that the energy use is not reasonable, and the comprehensive energy consumption quota is reduced by 5%; if the change trend of the important features of a building in the past several years is a decreasing trend, and the decreasing amplitude is greater than 20%, it is judged that the building actively responds to the energy saving and emission reduction policy, and the comprehensive energy consumption quota is increased by 5%.

7. A system for public building energy consumption quota based on multi-level classification and dynamic data correction, characterized in that, The method for determining the energy consumption quota of public buildings based on multi-level classification and dynamic data correction according to any one of claims 1-6 comprises a data collection and processing module, a data preprocessing module, a building type first-level classification module, a building type second-level classification module, an energy consumption quota module, a feature importance analysis module, and an energy consumption quota value correction module; The data collection and processing module is used to obtain the monthly and annual energy consumption parameters of different types of public buildings, including date parameters, load parameters and energy saving measures parameters; and a multi-feature data set is constructed based on the obtained energy consumption parameters; The data preprocessing module is used for data verification and cleaning, data feature screening, and data normalization processing of the multi-feature data set; The building type first-level classification module is used to classify and label the sample set according to the functionality and energy use attributes of public buildings; The building type second-level classification module is used to perform second-level classification on the results of the first-level logical and classification labels according to the K-means model, and logically explain and label the results of the second-level classification according to the feature information. The energy consumption quota module is configured to determine energy consumption quotas of public buildings of the same secondary type according to the quota level and ranking coupling method and the simulation quota method respectively, and obtain a comprehensive energy consumption quota by weighted averaging of the quotas obtained by the two methods. The feature importance analysis module is configured to perform annual energy consumption feature importance analysis of different types of public buildings based on SHAP value model according to energy consumption parameters, energy consumption prediction model and energy consumption prediction results, and obtain important features and Shapley value weights of the important features. The energy consumption quota value correction module is configured to perform linear regression on annual data averages of important features of each type of public building, and predict average values and change rates of the important features of each type of public building in future years according to the regression curve. The comprehensive energy consumption quota of each type of building in future years is obtained by weighting correction of the change rate of the important features and the Shapley value weights of the important features. The change rate in the past years obtained by linear regression on monthly data of important features of a single building is used to determine the energy use rationality and energy saving enthusiasm of the single building, and the comprehensive energy consumption quota of the single building is corrected.

8. A storage medium storing a program, characterized by comprising: The program is executed by the processor to implement the method for determining energy consumption quotas of public buildings based on multi-level classification and dynamic data correction according to any one of claims 1-6.

Images