Multi-source energy data analysis method and device, computer equipment and storage medium

Abstract

The invention relates to a multi-source energy data analysis method and device, computer equipment and a storage medium. The method comprises the following steps: acquiring multi-source energy data and a plurality of energy data analysis targets; performing fusion processing on the multi-source energy data to obtain multi-source energy fusion data; for each energy data analysis target, identifying a key variable matched with the energy data analysis target from the multi-source energy fusion data; for each energy data analysis target, determining an association relationship between a key variable matched with the energy data analysis target and each type of energy data in the multi-source energy fusion data, and screening basic energy data related to the energy data analysis target from the multi-source energy fusion data according to the association relationship; and constructing an energy data analysis model corresponding to each energy data analysis target according to the key variables and the basic energy data matched with each energy data analysis target. The method is beneficial to improving the accuracy of energy data analysis.

Description

Technical Field

[0001] This application relates to the field of energy ecological services, and in particular to a multi-source energy data processing method, apparatus, computer equipment, computer-readable storage medium, and computer program product. Background Technology

[0002] As energy systems become increasingly complex, the sources of energy data on both the grid and user sides are becoming more diverse. These data include, but are not limited to, smart meter data, sensor data, meteorological data, and user behavior data. These data are rich in type, varied in format, and massive in volume, covering all aspects of energy production, transmission, and consumption. Therefore, accurately analyzing the intrinsic relationships between different data sources based on the fusion of multiple data sources has become a pressing problem in the field of energy data processing and analysis.

[0003] However, current energy data processing and analysis methods have many shortcomings when dealing with multi-source heterogeneous data, resulting in low accuracy in energy data analysis. For example, it is difficult to effectively uncover the intrinsic connections between different data sources, and the complex nonlinear relationship between user behavior data and energy consumption is difficult to reveal through simple statistical analysis methods. Summary of the Invention

[0004] Therefore, it is necessary to provide a multi-source energy data analysis method, apparatus, computer equipment, computer-readable storage medium, and computer program product that can improve the accuracy of multi-source energy data processing, in response to the above-mentioned technical problems.

[0005] Firstly, this application provides a method for analyzing multi-source energy data, including:

[0006] Acquire multi-source energy data and achieve multiple energy data analysis objectives;

[0007] Multi-source energy data is fused to obtain multi-source energy fused data.

[0008] For each energy data analysis objective, key variables that match the energy data analysis objective are identified from multi-source energy fusion data;

[0009] For each energy data analysis objective, determine the correlation between the key variables matching the energy data analysis objective and each type of energy data in the multi-source energy fusion data. Based on the correlation, select the basic energy data related to the energy data analysis objective from the multi-source energy fusion data.

[0010] Based on the key variables and basic energy data that match each energy data analysis objective, construct the corresponding energy data analysis model for each energy data analysis objective.

[0011] Secondly, this application also provides a multi-source energy data analysis device, comprising:

[0012] The data acquisition module is used to acquire multi-source energy data and multiple energy data analysis targets;

[0013] The data fusion module is used to fuse multi-source energy data to obtain multi-source energy fused data;

[0014] The key variable identification module is used to identify key variables that match the energy data analysis objectives from multi-source energy fusion data for each energy data analysis objective.

[0015] The basic energy data identification module is used to determine the correlation between the key variables matching each energy data analysis target and each type of energy data in the multi-source energy fusion data for each energy data analysis target, and to filter out the basic energy data related to the energy data analysis target from the multi-source energy fusion data based on the correlation.

[0016] The data analysis model building module is used to construct energy data analysis models corresponding to each energy data analysis objective based on the key variables and basic energy data that match each energy data analysis objective.

[0017] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in any of the above embodiments of the multi-source energy data analysis method.

[0018] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps in any of the above embodiments of the multi-source energy data analysis method.

[0019] Fifthly, this application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps in any of the above embodiments of the multi-source energy data analysis method.

[0020] The aforementioned multi-source energy data analysis methods, devices, computer equipment, computer-readable storage media, and computer program products acquire multi-source energy data and multiple energy data analysis targets, fuse the multi-source energy data to obtain multi-source energy fusion data, identify key variables matching the energy data analysis target from the multi-source energy fusion data for each energy data analysis target, perform correlation analysis between the key variables matching the energy data analysis target and each type of energy data in the multi-source energy fusion data to determine the correlation relationship, and then, based on the correlation relationship, filter out basic energy data related to the energy data analysis target from the multi-source energy fusion data. This facilitates in-depth exploration of the intrinsic connections between different data sources and reveals the complex relationships hidden in the data. Finally, based on the key variables and basic energy data matching each energy data analysis target, construct energy data analysis models corresponding to each energy data analysis target. This allows for the use of different energy data analysis models for different energy data, effectively handling the complex nonlinear relationship between user behavior data and energy consumption, and improving the model's adaptability to energy data analysis. Attached Figure Description

[0021] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the description of the embodiments of this application or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 This is a diagram illustrating the application environment of a multi-source energy data analysis method in one embodiment.

[0023] Figure 2 This is a flowchart illustrating a multi-source energy data analysis method in one embodiment;

[0024] Figure 3 This is a flowchart illustrating a multi-source energy data analysis method in another embodiment;

[0025] Figure 4 This is a flowchart illustrating a multi-source energy data analysis method in yet another embodiment;

[0026] Figure 5 This is a structural block diagram of a multi-source energy data analysis device in one embodiment;

[0027] Figure 6 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0028] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0029] The multi-source energy data analysis method provided in this application embodiment can be applied to, for example... Figure 1 In the application environment shown, terminal 102 communicates with server 104 via a network. A data storage system can store the data that server 104 needs to process. The data storage system can be integrated onto server 104, or it can be located in the cloud or on another network server.

[0030] Specifically, the operator can upload multiple energy data analysis targets and collected multi-source energy data to server 104 via terminal 102, and then send multi-source energy data analysis messages to server 104 via terminal 102. Server 104 obtains multi-source energy data and multiple energy data analysis target data. Next, it performs fusion processing on the multi-source energy data to obtain multi-source energy fusion data. Then, for each energy data analysis target, it identifies key variables that match the energy data analysis target from the multi-source energy fusion data. For each energy data analysis target, it determines the correlation between the key variables that match the energy data analysis target and each type of energy data in the multi-source energy fusion data. Based on the correlation, it filters out basic energy data related to the energy data analysis target from the multi-source energy fusion data. Finally, based on the key variables and basic energy data that match each energy data analysis target, it constructs an energy data analysis model corresponding to each energy data analysis target.

[0031] The terminal 102 can be, but is not limited to, various personal computers, laptops, smartphones, tablets, IoT devices, and portable wearable devices. IoT devices can include smart speakers, smart TVs, smart air conditioners, smart in-vehicle systems, and projection devices. Portable wearable devices can include smartwatches, smart bracelets, and head-mounted displays. Head-mounted displays can be virtual reality (VR) devices, augmented reality (AR) devices, and smart glasses. The server 104 can be a standalone physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing cloud computing services.

[0032] In one exemplary embodiment, such as Figure 2 As shown, a multi-source energy data analysis method is provided, which can be applied to... Figure 1Taking server 104 as an example, the explanation includes the following steps (hereinafter referred to as S): S100 to S400. Wherein:

[0033] S100 acquires multi-source energy data and multiple energy data analysis targets.

[0034] The objectives of energy data analysis include, but are not limited to, energy demand forecasting and energy resource allocation optimization.

[0035] Multi-source energy data can include multi-source energy data from the power grid side, multi-source energy data from the user side, and environmental data.

[0036] The multi-source energy data on the grid side can include, but is not limited to, smart meter data, substation operation data, and distributed energy data. Smart meter data can include, but is not limited to, the power load (active power, reactive power), voltage, and current of grid equipment. This data can be collected periodically from the meter reading system by setting data acquisition scripts. Substation operation data can include, but is not limited to, the operating status (temperature, pressure, number of operations, etc.), bus voltage, line current, and power factor of substation equipment such as transformers, circuit breakers, and disconnectors. This data can be obtained through the substation's SCADA system (Supervisory and Data Acquisition System). Distributed energy data includes, but is not limited to, the power generation capacity and duration of solar and wind power generation equipment. This data can be obtained through an energy management system that integrates distributed power sources, energy storage, and data collection.

[0037] The multi-source energy data on the user side can include, but is not limited to, user smart meter data and smart home device data. User smart meter data can include electricity load (active power, reactive power), voltage, current, water consumption, gas consumption, and cumulative electricity consumption (active power, reactive power), etc., and can be obtained by an automatic meter reading system. Smart home device data can include the power, on / off time, and running time of different types of devices (air conditioners, water heaters, lighting, sockets, etc.), etc., and can be obtained by the communication modules built into the home devices uploading device data to the cloud platform and collecting it through the cloud platform.

[0038] Environmental data may include, but is not limited to, meteorological data and energy subsidy strategies. Meteorological data may include temperature, humidity, wind speed, sunshine duration, etc., and can be collected from open-source meteorological databases. Energy subsidy strategies may include, but are not limited to, subsidy strategies for various renewable energy sources and electricity price adjustment strategies, and can be obtained through industry regulatory platforms.

[0039] In practice, operators can collect data from smart meters, substations, and distributed energy sources on the power grid side through automatic meter reading systems, monitoring and data acquisition systems, and energy management systems. They can also collect user smart meter data and smart home device data through user-side automatic meter reading systems and cloud platforms. Furthermore, they can collect meteorological data and energy subsidy policies through open-source meteorological databases and industry regulatory platforms. The collected multi-source energy data undergoes data cleaning and preprocessing. The cleaned and preprocessed data is then uploaded to a server via a terminal. The server receives multi-source energy data including data from the power grid, user-side, and environmental data. Subsequently, the operator sends a multi-source energy data analysis message carrying multiple energy data analysis targets to the server via the terminal. The server receives and parses the message to obtain multiple energy data analysis targets.

[0040] S200 performs fusion processing on multi-source energy data to obtain multi-source energy fusion data.

[0041] In practice, the server performs data fusion processing on multi-source energy data according to a preset data fusion strategy to obtain multi-source energy fused data. The data fusion strategy may include, but is not limited to, weighted average methods and linear combination methods. Subsequently, the server aggregates the multi-source energy fused data obtained from the fusion process into a global energy data table and stores the global energy data table in the database. The global energy data table includes fields such as timestamp, data type, and data value.

[0042] In other implementations, a real-time data update mechanism can be established, whereby the server responds to the real-time data update command sent by the operator through the terminal, obtains updated multi-source energy data, and updates the global energy data table, so that the global energy data table can reflect the latest multi-source energy data and environmental data from the grid side and the user side in a timely manner.

[0043] S300 identifies key variables that match the energy data analysis objective from multi-source energy fusion data for each energy data analysis objective.

[0044] Among them, key variables represent the key variables that affect the objectives of energy data analysis.

[0045] In practical applications, operators can pre-determine the mapping relationship between each energy data analysis objective and the key variables affecting that objective, based on their analytical needs. For example, the energy data analysis objective might be energy demand forecasting, with associated key variables including the demand for various types of energy. In implementation, the server, based on the established mapping relationship, selects data types from multi-source energy fusion data that have a mapping relationship with the energy data analysis objective—that is, the key variables that match the energy data analysis objective—for each energy data analysis objective.

[0046] S400, for each energy data analysis objective, determines the correlation between the key variables matching the energy data analysis objective and each type of energy data in the multi-source energy fusion data, and based on the correlation, selects the basic energy data related to the energy data analysis objective from the multi-source energy data.

[0047] The basic energy data can be the data type required to construct the analytical model for the energy data analysis objective. For example, for the energy demand forecasting objective, the relevant basic energy data may include historical load data, temperature, and humidity. For the equipment fault diagnosis objective, the relevant basic energy data may include equipment type, load data, and temperature.

[0048] In practice, for each energy data analysis objective, the server can use a correlation analysis algorithm to perform correlation analysis on key variables matching the energy data analysis objective with each type of energy data in the multi-source energy fusion data. The correlation analysis results are then used to determine whether there is a correlation between the key variable and each energy data. Energy data that is correlated with the key variable is identified as basic energy data relevant to the energy data analysis objective. Correlation analysis algorithms include, but are not limited to, Pearson correlation coefficient and Spearman rank correlation coefficient.

[0049] For example, for energy demand forecasting, the Spearman rank correlation coefficients of key variables and each energy data in multi-source energy fusion data are calculated, and it is detected whether the Spearman rank correlation coefficients are greater than or equal to a preset Spearman rank correlation coefficient threshold. If the detection results indicate that the Spearman rank correlation coefficients of user electricity peak, average electricity consumption, and equipment operating status data are greater than or equal to the preset Spearman rank correlation coefficient threshold, it is determined that there is a correlation between the key variables and these energy data, and these energy data are identified as basic energy data.

[0050] S500 constructs energy data analysis models corresponding to each energy data analysis objective based on key variables and basic energy data that match the objectives of each energy data analysis.

[0051] In practical implementation, for each energy data analysis target, the server may call a pre-trained model matching the energy data analysis target, and iteratively train the model based on key variables and basic energy data matching the energy data analysis target until a preset training termination condition is reached, thus obtaining an energy data analysis model corresponding to that energy data analysis target. The preset training termination condition may be that the loss value is continuously less than a preset loss threshold within a preset number of iterations, at which point model training stops. In other implementations, after constructing the energy data analysis module corresponding to each energy data analysis target, the method further includes: for each energy data analysis target, obtaining key variables and basic energy data matching the energy data analysis target through a global energy data table; using the key variables and basic energy data as input, calling the energy data analysis model corresponding to the energy data analysis target to obtain the energy data analysis results for that energy data analysis target.

[0052] Key variables and basic energy data obtained from actual measurements that match the objectives of energy data analysis.

[0053] In other implementations, for each energy data analysis target, the server can automatically generate features using automated feature engineering tools based on key variables and basic energy data that match the energy data analysis target. From these generated features, the most helpful features for the analysis target are selected. Then, a pre-trained model matching the energy data analysis target is invoked, and the selected features are used to iteratively train the model until a preset training termination condition is met, resulting in an energy data analysis model corresponding to that energy data analysis target. The selection of the most helpful features from the generated features can be achieved by performing correlation analysis (such as Pearson correlation coefficient) between the generated features and the target variables of the energy analysis target, and identifying features that meet a preset threshold as target features (features most helpful to the analysis target).

[0054] In the aforementioned multi-source energy data analysis method, multi-source energy data and multiple energy data analysis targets are acquired and fused to obtain multi-source energy fused data. For each energy data analysis target, key variables matching the energy data analysis target are identified from the multi-source energy fused data. Next, correlation analysis is performed between the key variables matching the energy data analysis target and each type of energy data in the multi-source energy fused data to determine the correlation. Subsequently, based on the correlation, basic energy data related to the energy data analysis target is selected from the multi-source energy fused data. This facilitates in-depth exploration of the intrinsic connections between different data sources and reveals the complex relationships hidden in the data. Then, based on the key variables and basic energy data matching each energy data analysis target, an energy data analysis model corresponding to each energy data analysis target is constructed. This allows for the use of different energy data analysis models for different energy data, effectively handling the complex nonlinear relationship between user behavior data and energy consumption, and improving the model's adaptability to energy data analysis.

[0055] In an exemplary embodiment, the energy data analysis objectives include energy data analysis objectives on the user side, the grid side, and the environment side. Based on the key variables and basic energy data matching each energy data analysis objective, an energy data analysis model corresponding to each energy data analysis objective is constructed, such as... Figure 3 As shown, this includes S520 to S540:

[0056] S520: Based on the pre-defined mapping relationship between energy data analysis objectives of each dimension and multiple sub-energy data analysis model categories, determine the sub-energy data analysis model categories corresponding to the energy data analysis objectives of each dimension.

[0057] The energy data analysis objectives on the user side may include, but are not limited to, user energy consumption habit analysis and user behavior analysis of demand response. The energy analysis objectives on the grid side may include, but are not limited to, energy consumption forecasting, energy demand forecasting, energy allocation optimization, energy optimization scheduling, power equipment fault identification, and grid system stability early warning. The energy data analysis on the environment side may include, but is not limited to, carbon emission analysis, renewable energy integration, energy price forecasting, and energy supply and demand trend analysis. The sub-energy data analysis model categories may be model categories corresponding to the sub-energy data analysis objectives under the overall energy data analysis objectives.

[0058] In practical applications, to achieve more refined energy data analysis, the energy data analysis objectives of each dimension can be refined in advance, and sub-energy data analysis objectives under each dimension can be refined. Then, the sub-energy data analysis model categories corresponding to each sub-energy data analysis objective are determined, and the mapping relationship between each dimension of energy data analysis objectives and multiple sub-energy data analysis model categories is constructed.

[0059] For example, the energy data analysis objectives at the user-side level include user energy consumption habit analysis and user behavior analysis of demand response. The energy data analysis model corresponding to the user-side energy analysis objectives can be a user behavior analysis model. Further refining the user-side energy data analysis objectives, the sub-energy data analysis model categories corresponding to the refined sub-energy data analysis objectives can include: a user classification sub-model: classifying users according to their energy consumption habits and behavioral patterns to provide a basis for personalized services; and a demand response analysis sub-model: analyzing user response behavior to price signals, incentives, etc., to optimize demand response strategies.

[0060] The objectives of energy data analysis for the grid side are as follows: (1) Energy demand forecasting analysis for predicting future energy consumption. The corresponding energy demand forecasting model includes the following sub-energy data analysis model categories: Short-term forecasting sub-model: predicts energy demand in the next few hours or days for grid dispatching and equipment operation planning. Long-term forecasting sub-model: predicts energy demand in the next few years for energy infrastructure planning and investment decisions. (2) Energy optimization scheduling analysis for optimizing energy allocation and use. The corresponding energy optimization scheduling model includes the following sub-energy data analysis model categories: Multi-energy system optimization sub-model: optimizes the allocation and use of various energy sources such as electricity, natural gas, and heat to improve energy utilization efficiency; Cost optimization sub-model: reduces energy procurement and operating costs while meeting energy demand. (3) Fault diagnosis and early warning analysis for identifying equipment faults or abnormal situations. The corresponding equipment fault diagnosis and early warning analysis model includes the following sub-energy data analysis model categories: Equipment fault diagnosis sub-model: identifies early signs of equipment faults through sensor data and historical data to reduce downtime; System stability early warning sub-model: monitors the stability of the power grid or energy system and provides early warning of potential faults or abnormal situations. (4) The energy efficiency improvement analysis target corresponds to the energy efficiency improvement analysis model, including the following sub-energy data analysis model categories: Energy saving potential analysis sub-model: identify the links of energy waste, propose energy saving measures and improvement suggestions; Energy efficiency assessment sub-model: assess the energy efficiency level of the energy system and compare it with industry standards or best practices.

[0061] Energy data analysis on the environmental side can include, but is not limited to, carbon emission analysis, renewable energy integration, energy price forecasting, and energy supply and demand trend analysis.

[0062] For the environmental side, the energy data analysis objectives are: (1) Energy market analysis objectives correspond to energy market analysis models, which include the following sub-energy data analysis model categories: Price prediction sub-model: predicts energy market price fluctuations and provides decision support for energy trading; Market trend analysis sub-model: analyzes the supply and demand relationship, policy impact, etc. of the energy market and provides a basis for strategic planning. (2) Environmental impact assessment analysis objectives correspond to environmental impact assessment analysis models, which include the following sub-energy data analysis model categories: Carbon emission analysis sub-model: assesses the carbon emission level of the energy system and proposes emission reduction measures; Renewable energy integration sub-model: analyzes the impact of renewable energy (such as solar and wind power) access on the power grid and energy system.

[0063] In practice, the server determines the sub-energy data analysis model category corresponding to each dimension of energy data analysis objective based on the pre-built mapping relationship between the energy data analysis objectives of each dimension and multiple sub-energy data analysis model categories.

[0064] S520 constructs a sub-energy data analysis model corresponding to each dimension of energy data analysis objectives, based on the key variables, basic energy data, and sub-energy data analysis model categories that match the energy data analysis objectives.

[0065] In practical implementation, for each dimension of energy data analysis objectives, the server can call a pre-trained model that matches the sub-energy data analysis type corresponding to the energy data analysis objective. The model is then iteratively trained based on key variables and basic energy data matching the energy data analysis objective until a preset training termination condition is met, resulting in a sub-energy data analysis model corresponding to that energy data analysis objective. The preset training termination condition can be that the loss value continuously falls below a preset loss threshold within a preset number of iterations, at which point model training stops.

[0066] In this embodiment, a sub-energy data analysis model corresponding to the energy data analysis target is constructed through a preset mapping relationship. The energy data analysis target is then analyzed based on the sub-energy data analysis target, which helps to improve the targeting and effectiveness of energy data analysis, as well as the analysis efficiency.

[0067] In one exemplary embodiment, based on a preset data correlation analysis strategy, the correlation between key variables and each type of energy data in the multi-source energy fusion data is determined, such as... Figure 4 As shown, it includes at least one of the following methods S422 to S428:

[0068] S422, based on a preset correlation analysis algorithm, determines whether there is a linear correlation between key variables and each type of energy data in the multi-source energy fusion data.

[0069] In practice, the server can use a preset correlation analysis algorithm to perform correlation analysis on the key variables and each type of energy data in the multi-source energy fusion data, obtain the correlation analysis results, and check whether the correlation analysis results meet the preset correlation analysis requirements to determine whether there is a linear correlation between the key variables and each type of energy data in the multi-source energy fusion data. The correlation analysis algorithm may include, but is not limited to, Pearson correlation coefficient and Spearman rank correlation coefficient. For example, the Pearson correlation coefficient between the key variables and each type of energy data in the multi-source energy fusion data is determined. The Pearson correlation coefficient is a statistic that measures the degree of linear correlation between two variables, and its calculation formula is:

[0070]

[0071] in, and These are the observed values ​​of the two variables. and These are their means. r=1: perfectly positive linear correlation. r=-1: perfectly negative linear correlation. r=0: no linear correlation. The closer the absolute value of r is to 1, the stronger the linear correlation; the closer it is to 0, the weaker the linear correlation. Therefore, the correlation analysis requirement for the Pearson correlation coefficient can be: if r is greater than or equal to a preset Pearson correlation coefficient threshold, then the two variables are considered to be linearly correlated. For example, the Spearman rank correlation coefficient between the key variable and each type of energy data in the multi-source energy fusion data is determined based on the following formula:

[0072]

[0073] in, The Spearman rank correlation coefficient is the difference in rank between two variables. It is a statistic that measures the degree of monotonic correlation between two variables and is applicable to nonlinear relationships or uneven data distributions. ρ=1: perfectly positive monotonic correlation. ρ=-1: perfectly negative monotonic correlation. ρ=0: no monotonic correlation. The closer the absolute value of ρ is to 1, the stronger the monotonic correlation; the closer it is to 0, the weaker the monotonic correlation. Therefore, the correlation analysis requirement for the Spearman rank correlation coefficient is that ρ is greater than or equal to a preset Pearson correlation coefficient threshold, in which case the two variables are considered to be linearly correlated.

[0074] In other embodiments, based on a preset data correlation analysis strategy, determining the correlation between key variables and each type of energy data in the multi-source energy fusion data further includes: constructing a scatter plot of key variables and each type of energy data in the multi-source energy fusion data; if the scatter plot shows a linear distribution, determining that the key variables and energy data are linearly correlated; if the scatter plot shows a curved distribution, determining that the key variables and energy data are non-linearly correlated.

[0075] In other embodiments, the determination of the correlation between key variables and each type of energy data in multi-source energy fusion data based on a preset data correlation analysis strategy further includes: determining the goodness of fit between key variables and each type of energy data in multi-source energy fusion data based on a trained regression model; and determining that the key variables and energy data are linearly correlated if the goodness of fit is greater than or equal to a preset goodness of fit threshold.

[0076] In other embodiments, based on a preset data correlation analysis strategy, the correlation between key variables and each type of energy data in multi-source energy fusion data is determined. This also includes: based on a constructed multiple linear regression model, analyzing the impact of each type of energy data in multi-source energy fusion data on key variables, obtaining regression coefficients between each type in multi-source energy fusion data, whereby the regression coefficients characterize the degree of influence of energy data on key variables. Subsequently, the regression coefficients are compared with a preset regression coefficient threshold. If the regression coefficient between two variables is greater than or equal to the preset regression coefficient threshold, then the two variables are determined to be linearly correlated.

[0077] S424, based on a causal analysis algorithm, determines whether there is a causal relationship between key variables and each type of energy data in multi-source energy fusion data.

[0078] In practice, the server can use a pre-defined causal analysis algorithm to perform causal analysis on each type of energy data in the multi-source energy fusion data, obtaining causal analysis results. The results are then checked to determine whether they meet pre-defined causal analysis requirements, thus determining whether there is a causal correlation between the key variables and each type of energy data in the multi-source energy fusion data. Causal analysis algorithms include, but are not limited to, Granger causality tests, instrumental variable methods, and differences-in-differences methods. The pre-defined causal analysis requirement for the Granger causality test is that the p-value between two variables is less than a pre-set significance level (e.g., 0.05), indicating a causal relationship between the two variables, i.e., a causal correlation. For example, the Granger causality test can be used to analyze whether meteorological conditions lead to significant changes in energy consumption, or whether user behavior affects energy usage patterns.

[0079] S426, determine the mutual information between the key variables and each type of energy data in the multi-source energy fusion data, and based on the mutual information, determine whether there is an information correlation between the key variables and each type of energy data in the multi-source energy fusion data.

[0080] Mutual information is used to measure the degree of information sharing between two variables and can be used to detect dependencies between variables.

[0081] In practice, the server can determine the mutual information between the key variable and each type of energy data in the multi-source energy fusion data based on the mutual information calculation formula, and detect whether the mutual information meets the preset mutual information threshold. If the mutual information is greater than or equal to the preset mutual information threshold, it can be determined whether the key variable and each type of energy data in the multi-source energy fusion data are informationally related.

[0082] S428, based on a preset spatial correlation analysis algorithm, determines whether there is a spatial correlation between key variables and each type of energy data in the multi-source energy fusion data.

[0083] Spatial correlation analysis algorithms include, but are not limited to, Moran coefficient and local Moran index.

[0084] In practical implementation, the server can determine the spatial correlation coefficient between key variables and each type of energy data in the multi-source energy fusion data based on a preset spatial correlation analysis algorithm, and detect whether the spatial correlation coefficient is greater than or equal to a preset spatial correlation coefficient threshold to determine whether there is a spatial correlation between the key variables and each type of energy data in the multi-source energy fusion data. For example, the Moran coefficient between the key variables and each type of energy data in the multi-source energy fusion data can be determined based on the following formula:

[0085]

[0086] Where N is the number of observation points, x i and x j These are observed values. It is the mean of the observed values, w ij It is a spatial weight matrix, representing the spatial relationship between observation points i and j, and W is the sum of all spatial weights.

[0087] Subsequently, the Moran coefficient is compared with a preset Moran coefficient threshold. If the Moran coefficient between the two variables is greater than or equal to the preset Moran coefficient threshold, the two variables are determined to be spatially correlated.

[0088] In this embodiment, a multi-dimensional correlation analysis is performed on the key variables and each type of energy data in the multi-source energy fusion data. This helps to improve the accuracy of the correlation determination, thereby improving the accuracy of the energy data analysis model construction.

[0089] In other embodiments, determining the association between key variables and each type of energy data in the multi-source energy fusion data based on a preset data association analysis strategy further includes: determining the dependency relationship between key variables and each type of energy data in the multi-source energy fusion data based on a trained deep neural network. The deep neural network includes, but is not limited to, recurrent neural networks and their variants (such as long short-term memory networks and GRUs), used to capture temporal dependencies between data.

[0090] In one exemplary embodiment, based on the correlation between key variables and each energy data point, basic energy data relevant to the energy data analysis objectives are filtered from multi-source energy fusion data, including:

[0091] For each energy data point, if there is at least one correlation between the energy data and the key variables, namely linear correlation, causal correlation, informational correlation, and spatial correlation, the energy data is identified as the basic energy data associated with the energy data analysis objective.

[0092] In other embodiments, basic energy data related to the energy data analysis target are screened from multi-source energy fusion data based on the correlation between key variables and each energy data. This further includes: for each energy data, if there is at least one correlation between the energy data and the key variables, such as linear correlation, causal correlation, information correlation, dependency relationship and spatial correlation, the energy data is determined to be basic energy data related to the energy data analysis target.

[0093] In this embodiment, multi-dimensional correlation analysis helps improve the accuracy of correlation determination, thereby improving the accuracy of energy data analysis model construction.

[0094] In an exemplary embodiment, after constructing the energy data analysis model corresponding to each energy data analysis objective, the method further includes steps S620 to S680:

[0095] S620, respectively acquires energy data test sets that match each energy data analysis model.

[0096] The energy data test set for each energy data analysis model can include key variables and basic energy data that match the energy data analysis objectives.

[0097] In practice, the server can determine the energy data analysis target corresponding to the energy data analysis model, obtain key variables and basic energy data in historical time periods that match the energy data analysis target, and construct an energy data test set.

[0098] S640 inputs the energy data test set into the energy data analysis model to obtain the data analysis results.

[0099] In practice, the server, based on model evaluation methods, determines the corresponding energy data analysis model for each energy data analysis objective. The energy data test set corresponding to the energy data analysis model is then input into the model. The model analyzes the energy data analysis objective and outputs data analysis results corresponding to that objective. Model evaluation methods include, but are not limited to, hold-out, cross-validation, and bootstrapping.

[0100] S660 determines the model performance evaluation index values ​​based on the data analysis results.

[0101] In practice, different types of energy data analysis models correspond to different model performance evaluation metrics. For each energy data analysis model, the server determines the corresponding model performance evaluation metrics and, based on the data analysis results, determines the values ​​of these metrics. For example, for regression models, performance evaluation metrics may include, but are not limited to, mean squared error, mean absolute error, root mean square error, and coefficient of determination (R²). For classification models, performance evaluation metrics may include, but are not limited to, accuracy, precision, recall, and F1 score. For clustering models, performance evaluation metrics may include, but are not limited to, silhouette coefficient (measuring the quality of clustering) and Davis-Bundling index (measuring the separation and compactness of clusters).

[0102] S680 adjusts the hyperparameters of the energy data analysis model when the model performance evaluation index, which characterizes the data analysis performance of the energy data analysis model, does not meet the preset data analysis performance requirements.

[0103] In practical implementation, different energy data analysis models have different performance evaluation metrics for their model types, and these metrics have different performance requirements. Thresholds for these metrics can be pre-defined. The server compares the model performance evaluation metrics with their corresponding thresholds. If a metric meets the threshold, the energy data analysis model is deemed to meet the requirements; otherwise, it is deemed not to. For example, for clustering models, the performance requirement for the silhouette coefficient could be that the silhouette coefficient is greater than or equal to a preset threshold. Adjusting the hyperparameters of the energy data analysis model can be done by exhaustively searching a predefined hyperparameter grid to find the optimal combination of hyperparameters and updating the hyperparameters accordingly; it can also involve random sampling in the hyperparameter space to search for the optimal combination of hyperparameters and updating the hyperparameters accordingly; or it can be done by dynamically adjusting the search direction using Bayesian methods to efficiently find the optimal hyperparameters and updating the hyperparameters accordingly.

[0104] In other implementations, if the model performance evaluation index, which characterizes the data analysis performance of the energy data analysis model, does not meet the preset data analysis performance requirements, important features can be selected through methods such as correlation analysis and LightGBM (Light Gradient Boosting Machine) regression algorithm, or new features can be extracted through methods such as principal component analysis, and the model can be retrained and its performance evaluated. Alternatively, ensemble methods such as Bagging, Boosting, or Stacking can be used to improve the stability and performance of the model.

[0105] In this embodiment, the performance of the constructed analysis model is verified using the acquired data, and the model is optimized based on the verification results to improve the model's analysis performance.

[0106] In an exemplary embodiment, after acquiring multi-source energy data, the method further includes: preprocessing the multi-source energy data, wherein the preprocessing includes at least one of noise removal, outlier handling, and missing value imputation.

[0107] In practice, the server performs missing value imputation based on the data type of the multi-source energy data: for time-series data, missing value imputation may include, but is not limited to, forward imputation, backward imputation, or interpolation imputation; for non-time-series data, missing value imputation may include, but is not limited to, mean imputation or median imputation. Subsequently, outlier handling is performed on the multi-source energy data, which can be based on the degree of impact of outlier data. For example, if an outlier is caused by equipment failure or data transmission error, the data point can be directly deleted. Afterward, noise removal is performed on the multi-source energy data, which may include, but is not limited to, using filtering methods (such as low-pass filters and high-pass filters) to remove noise components from the multi-source energy data.

[0108] Multi-source energy data is fused to obtain multi-source energy fused data, including: fusing pre-processed multi-source energy data to obtain multi-source energy fused data.

[0109] In specific implementation, the preprocessed multi-source energy data is fused to obtain the implementation method of multi-source energy fusion data. The implementation method of multi-source energy fusion data is the same as that in the above embodiment, and will not be repeated here.

[0110] To provide a clearer explanation of the multi-source energy data analysis method provided in this application, a specific embodiment is described below, which includes the following steps:

[0111] S1 acquires multi-source energy data and multiple energy data analysis targets.

[0112] S2, preprocessing multi-source energy data, including at least one of noise removal, outlier handling, and missing value imputation.

[0113] S3 performs fusion processing on the preprocessed multi-source energy data to obtain multi-source energy fusion data.

[0114] S4 identifies key variables that match the energy data analysis objective from multi-source energy fusion data for each energy data analysis objective.

[0115] S5. For each energy data analysis target, perform the following steps: Based on a preset correlation analysis algorithm, determine whether there is a linear correlation between the key variable and each type of energy data in the multi-source energy fusion data; based on a causal analysis algorithm, determine whether there is a causal correlation between the key variable and each type of energy data in the multi-source energy fusion data; determine the mutual information between the key variable and each type of energy data in the multi-source energy fusion data; based on the mutual information, determine whether there is an informational correlation between the key variable and each type of energy data in the multi-source energy fusion data; based on a preset spatial correlation analysis algorithm, determine whether there is a spatial correlation between the key variable and each type of energy data in the multi-source energy fusion data; if there is at least one correlation relationship among linear correlation, causal correlation, informational correlation, and spatial correlation between the energy data and the key variable, determine the energy data as the basic energy data associated with the energy data analysis target.

[0116] S6. Based on the pre-defined mapping relationship between the energy data analysis objectives of each dimension and multiple sub-energy data analysis model categories, determine the sub-energy data analysis model categories corresponding to the energy data analysis objectives of each dimension. For each dimension of energy data analysis objectives, construct the sub-energy data analysis model corresponding to the energy data analysis objectives based on the key variables, basic energy data, and sub-energy data analysis model categories that match the energy data analysis objectives.

[0117] S7. Obtain energy data test sets that match each energy data analysis model, input the energy data test sets into the energy data analysis model, obtain data analysis results, determine the model performance evaluation index value based on the data analysis results, and adjust the hyperparameters of the energy data analysis model if the model performance evaluation index characterizes the data analysis performance of the energy data analysis model and does not meet the preset data analysis performance requirements.

[0118] It should be understood that although the steps in the flowcharts of the above embodiments are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the above embodiments may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0119] In one exemplary embodiment, such as Figure 5As shown, a multi-source energy data analysis device 600 is provided, including: a data acquisition module 610, a data fusion module 620, a key variable identification module 630, a basic energy data identification module 640, and a data analysis model construction module 650, wherein:

[0120] Data acquisition module 610 is used to acquire multi-source energy data and multiple energy data analysis targets;

[0121] The data fusion module 620 is used to fuse multi-source energy data to obtain multi-source energy fusion data;

[0122] The key variable identification module 630 is used to identify key variables that match the energy data analysis target from multi-source energy fusion data for each energy data analysis target.

[0123] The basic energy data identification module 640 is used to determine the correlation between the key variables matching the energy data analysis target and each type of energy data in the multi-source energy fusion data for each energy data analysis target, and to filter out the basic energy data related to the energy data analysis target from the multi-source energy fusion data based on the correlation.

[0124] The data analysis model building module 650 is used to build energy data analysis models corresponding to each energy data analysis objective based on the key variables and basic energy data that match each energy data analysis objective.

[0125] In an exemplary embodiment, the data analysis model building module 650 is further configured to determine the sub-energy data analysis model category corresponding to each dimension of energy data analysis target based on the preset mapping relationship between each dimension of energy data analysis target and multiple sub-energy data analysis model categories; and for each dimension of energy data analysis target, construct the sub-energy data analysis model corresponding to the energy data analysis target based on the key variables, basic energy data and sub-energy data analysis model categories that match the energy data analysis target.

[0126] In an exemplary embodiment, the basic energy data identification module 640 is further configured to: determine whether there is a linear correlation between the key variable and each type of energy data in the multi-source energy fusion data based on a preset correlation analysis algorithm; determine whether there is a causal correlation between the key variable and each type of energy data in the multi-source energy fusion data based on a causal analysis algorithm; determine the mutual information between the key variable and each type of energy data in the multi-source energy fusion data, and determine whether there is an informational correlation between the key variable and each type of energy data in the multi-source energy fusion data based on the mutual information; and determine whether there is a spatial correlation between the key variable and each type of energy data in the multi-source energy fusion data based on a preset spatial correlation analysis algorithm.

[0127] In an exemplary embodiment, the basic energy data identification module 640 is further configured to, for each energy data, determine that the energy data is basic energy data associated with the energy data analysis target if there is at least one correlation relationship among linear correlation, causal correlation, information correlation and spatial correlation between the energy data and the key variable.

[0128] In one exemplary embodiment, the multi-source energy data analysis device 600 further includes a model performance verification module 660 and a hyperparameter adjustment module 670:

[0129] The data acquisition module 610 is also used to acquire energy data test sets that match each energy data analysis model;

[0130] The model performance verification module 660 is used to input the energy data test set into the energy data analysis model to obtain the data analysis results; and to determine the model performance evaluation index value based on the data analysis results.

[0131] The hyperparameter adjustment module 670 is used to adjust the hyperparameters of the energy data analysis model when the model performance evaluation index characterizing the data analysis performance of the energy data analysis model does not meet the preset data analysis performance requirements.

[0132] In an exemplary embodiment, the multi-source energy data analysis device 600 further includes a data preprocessing module 680 for preprocessing the multi-source energy data, wherein the preprocessing includes at least one of noise removal, outlier handling, and missing value imputation.

[0133] The data fusion module 620 is also used to fuse the preprocessed multi-source energy data to obtain multi-source energy fusion data.

[0134] In one exemplary embodiment, the multi-source energy data analysis device 600 further includes a data analysis module 690:

[0135] The data acquisition module 610 is also used to acquire key variables and basic energy data that match the energy data analysis target through a global energy data table for each energy data analysis target.

[0136] The data analysis module 690 is used to take key variables and basic energy data as input, call the energy data analysis model corresponding to the energy data analysis target, and obtain energy data analysis results for the energy data analysis target.

[0137] Each module in the aforementioned multi-source energy data analysis device 600 can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the operations corresponding to each module.

[0138] In one exemplary embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 6 As shown, this computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operating system and computer programs stored in the non-volatile storage media. The database stores data. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communicating with external terminals via a network. When executed by the processor, the computer program implements a multi-source energy data analysis method.

[0139] Those skilled in the art will understand that Figure 6 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0140] In one exemplary embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in any of the above embodiments of the multi-source energy data analysis method.

[0141] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored, which, when executed by a processor, implements the steps in any of the above embodiments of the multi-source energy data analysis method.

[0142] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in any of the above embodiments of the multi-source energy data analysis method.

[0143] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0144] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, artificial intelligence (AI) processors, etc., and are not limited to these.

[0145] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0146] The above-described embodiments merely represent several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present application. It should be noted that a person of ordinary skill in the art may make various modifications and improvements without departing from the spirit of the present application, and these modifications and improvements fall within the scope of protection of the present application. Therefore, the scope of protection of the present application shall be determined by the appended claims.

Claims

1. A multi-source energy data analysis method, characterized in that, The method includes: Acquire multi-source energy data and achieve multiple energy data analysis objectives; The multi-source energy data is fused to obtain multi-source energy fused data; For each of the energy data analysis objectives, key variables that match the energy data analysis objective are identified from the multi-source energy fusion data; For each energy data analysis objective, determine the correlation between the key variables matching the energy data analysis objective and each type of energy data in the multi-source energy fusion data, and based on the correlation, filter out the basic energy data related to the energy data analysis objective from the multi-source energy fusion data; Based on the key variables and basic energy data that match each of the stated energy data analysis objectives, construct the energy data analysis model corresponding to each stated energy data analysis objective.

2. The method according to claim 1, characterized in that, The energy data analysis objectives include energy data analysis objectives from the user side, the grid side, and the environment side; The step of constructing an energy data analysis model corresponding to each energy data analysis objective based on key variables and basic energy data matching each energy data analysis objective includes: Based on the pre-defined mapping relationship between the energy data analysis objectives of each dimension and multiple sub-energy data analysis model categories, determine the sub-energy data analysis model categories corresponding to the energy data analysis objectives of each dimension; For each dimension of energy data analysis objectives, a sub-energy data analysis model is constructed based on the key variables, basic energy data, and sub-energy data analysis model categories that match the energy data analysis objectives.

3. The method according to claim 1, characterized in that, The predetermined data correlation analysis strategy determines the correlation between the key variables and each type of energy data in the multi-source energy fusion data, including at least one of the following: Based on a preset correlation analysis algorithm, it is determined whether the key variable is linearly correlated with each type of energy data in the multi-source energy fusion data; Based on the causal analysis algorithm, determine whether there is a causal relationship between the key variable and each type of energy data in the multi-source energy fusion data; Determine the mutual information between the key variable and each type of energy data in the multi-source energy fusion data, and based on the mutual information, determine whether there is an information correlation between the key variable and each type of energy data in the multi-source energy fusion data; Based on a preset spatial correlation analysis algorithm, it is determined whether there is a spatial correlation between the key variable and each type of energy data in the multi-source energy fusion data.

4. The method according to claim 3, characterized in that, The step of filtering basic energy data related to the energy data analysis objective from the multi-source energy fusion data based on the correlation between the key variables and each energy data point includes: For each of the energy data, if there is at least one correlation among linear correlation, causal correlation, information correlation, and spatial correlation between the energy data and the key variable, the energy data is determined to be basic energy data associated with the energy data analysis objective.

5. The method according to claim 1, characterized in that, After constructing the energy data analysis model corresponding to each of the energy data analysis objectives, the method further includes: Obtain energy data test sets that match each of the described energy data analysis models; The energy data test set is input into the energy data analysis model to obtain the data analysis results; Based on the data analysis results, determine the model performance evaluation index values; If the data analysis performance of the energy data analysis model does not meet the preset data analysis performance requirements as indicated by the model performance evaluation index, the hyperparameters of the energy data analysis model shall be adjusted.

6. The method according to any one of claims 1 to 5, characterized in that, After acquiring multi-source energy data, the method further includes: The multi-source energy data is preprocessed, and the preprocessing includes at least one of noise removal, outlier handling, and missing value imputation. The process of fusing the multi-source energy data to obtain multi-source energy fused data includes: The preprocessed multi-source energy data is fused to obtain multi-source energy fused data.

7. A multi-source energy data analysis device, characterized in that, The device includes: The data acquisition module is used to acquire multi-source energy data and multiple energy data analysis targets; The data fusion module is used to fuse the multi-source energy data to obtain multi-source energy fused data; The key variable identification module is used to identify key variables that match the energy data analysis target from the multi-source energy fusion data for each energy data analysis target. The basic energy data identification module is used to determine the correlation between the key variables matching the energy data analysis target and each type of energy data in the multi-source energy fusion data for each energy data analysis target, and to filter out basic energy data related to the energy data analysis target from the multi-source energy fusion data based on the correlation. The data analysis model building module is used to construct energy data analysis models corresponding to each energy data analysis objective based on the key variables and basic energy data that match each energy data analysis objective.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

Images