Green energy generation prediction method and related device based on power big data

Abstract

The present invention belongs to the technical field of power system automation, and specifically relates to a method and related device for predicting green energy power generation under power big data. The method includes: obtaining energy production data, meteorological data and economic indicator data in the power big data platform; extracting energy characteristics, time characteristics, spatial characteristics and climate characteristics from heterogeneous data; constructing a knowledge graph of entities and relationships based on the extracted characteristics; entities include time entities, location entities, climate entities and energy entities, and relationships include: time-energy relationships, climate-energy relationships and location-energy relationships; converting entities, relationships and attributes in the knowledge graph into graph structure data, and generating node vector representations; training a green energy analysis model based on the node vector representation, and using the green energy analysis model to predict future energy production; wherein the green energy analysis model includes a GCN layer and a GAT layer.

Description

Technical Field

[0001] The present invention belongs to the technical field of power system automation, and specifically relates to a green energy power generation prediction method and related devices under power big data. Background Art

[0002] Traditional energy, such as coal, oil and natural gas, is a limited resource and releases a large amount of greenhouse gases such as carbon dioxide during use. The development of green energy can reduce dependence on these fossil fuels, reduce demand for limited resources, and reduce greenhouse gas emissions, which will help address climate change issues.

[0003] The forecast analysis of green energy production is a typical time series forecasting problem. Among the time series forecasting methods, there are data-based or model-based forecasting methods, as well as hybrid methods that combine the two. However, when traditional solutions apply complex models such as polynomial regression, the interpretability of the model may be affected, and it is difficult to understand the impact of model parameters on the results, resulting in low accuracy and reliability in predicting green energy production and consumption. Summary of the invention

[0004] The purpose of the present invention is to provide a green energy power generation prediction method and related devices under power big data to solve the problems of low accuracy and reliability in predicting green energy production and consumption in the prior art.

[0005] In order to achieve the above object, the present invention adopts the following technical solution:

[0006] In a first aspect, the present invention provides a method for predicting green energy generation under power big data, comprising the following steps:

[0007] Acquire heterogeneous data in the power big data platform, including energy production data, meteorological data, and economic indicator data;

[0008] Extracting features from the heterogeneous data, including energy features, time features, spatial features, and climate features;

[0009] Based on the extracted features, a knowledge graph of entities and relationships is constructed; wherein the entities include time entities, location entities, climate entities and energy entities, and the relationships include: time-energy relationships, climate-energy relationships and location-energy relationships; entities, relationships and attributes in the knowledge graph are converted into graph structure data, and node vector representations are generated based on the graph structure data;

[0010] Based on the node vector representation, a pre-trained green energy analysis model is used to predict future energy production; wherein the green energy analysis model includes a GCN layer and a GAT layer.

[0011] Furthermore, the entities, relationships and attributes in the knowledge graph are converted into graph structure data, and node vector representations are generated based on the graph structure data, including the following steps:

[0012] The entities in the knowledge graph are taken as nodes of the graph, and the relationships are taken as edges of the graph. A unique identifier is assigned to each entity and relationship. The attributes of the entity are taken as node features, and the graph is constructed using the graph database Neo4j to obtain graph structure data.

[0013] Based on the Node2Vec algorithm, multiple random walks are performed from each node of the graph, and each random walk generates a node sequence; wherein the random walk is controlled by the return parameter p and the entry and exit parameter q; after the random walk is completed, the random walk sequence and the node sequence corresponding to each random walk are obtained;

[0014] The Word2Vec algorithm is used to convert the node sequence in the random walk sequence into a node vector representation.

[0015] Furthermore, the Word2Vec algorithm is used to convert the node sequence in the random walk sequence into a node vector representation, including:

[0016] In the Word2Vec algorithm, the node sequence is used as the input data of the Skip-gram model. The Skip-gram model learns the vector representation of the node by predicting the neighbor nodes in the node sequence.

[0017] Furthermore, the green energy analysis model includes a GCN layer and a GAT layer, wherein:

[0018] The GCN layer is used to utilize the spatial structure information of the graph and propagate features through the adjacency matrix and degree matrix;

[0019] The GAT layer captures the importance of nodes by learning the attention weights between nodes, so that the green energy analysis model pays attention to neighbor nodes whose importance meets the preset requirements. It combines the learned node vector representation as the initial feature and uses the attention mechanism to aggregate the information of neighbor nodes to obtain the trained green energy analysis model.

[0020] Furthermore, after the step of using the trained green energy analysis model to predict future energy production, it also includes:

[0021] Using green energy analysis models, identify the main factors affecting energy production;

[0022] Combining the predicted results of energy production and major factors, formulate power generation plans or adjust power generation equipment parameters.

[0023] Furthermore, after the step of obtaining heterogeneous data in the power big data platform, including energy production data, meteorological data and economic indicator data, it also includes:

[0024] The heterogeneous data are integrated and cleaned to unify the data formats and standards.

[0025] In a second aspect, the present invention provides a green energy generation prediction device under power big data, comprising:

[0026] Data acquisition module, used to acquire heterogeneous data in the power big data platform, including energy production data, meteorological data and economic indicator data;

[0027] A feature extraction module, used to extract features from the heterogeneous data, including energy features, time features, space features and climate features;

[0028] A graph construction module is used to construct a knowledge graph of entities and relationships based on the extracted features; wherein the entities include time entities, location entities, climate entities and energy entities, and the relationships include: time-energy relationships, climate-energy relationships and location-energy relationships; converting entities, relationships and attributes in the knowledge graph into graph structure data, and generating node vector representations based on the graph structure data;

[0029] The production prediction module is used to predict future energy production based on the node vector representation using a pre-trained green energy analysis model; wherein the green energy analysis model includes a GCN layer and a GAT layer.

[0030] Furthermore, in the graph construction module, entities, relationships, and attributes in the knowledge graph are converted into graph structure data, and node vector representations are generated based on the graph structure data, including the following:

[0031] The entities in the knowledge graph are taken as nodes of the graph, and the relationships are taken as edges of the graph. A unique identifier is assigned to each entity and relationship. The attributes of the entity are taken as node features, and the graph is constructed using the graph database Neo4j to obtain graph structure data.

[0032] Based on the Node2Vec algorithm, multiple random walks are performed from each node of the graph, and each random walk generates a node sequence; wherein the random walk is controlled by the return parameter p and the entry and exit parameter q; after the random walk is completed, the random walk sequence and the node sequence corresponding to each random walk are obtained;

[0033] The Word2Vec algorithm is used to convert the node sequence in the random walk sequence into a node vector representation.

[0034] Furthermore, the Word2Vec algorithm is used to convert the node sequence in the random walk sequence into a node vector representation, including:

[0035] In the Word2Vec algorithm, the node sequence is used as the input data of the Skip-gram model. The Skip-gram model learns the vector representation of the node by predicting the neighbor nodes in the node sequence.

[0036] In a third aspect of the present invention, an electronic device is provided, comprising a processor and a memory, wherein the processor is used to execute a computer program stored in the memory to implement the green energy power generation prediction method under the power big data as described above.

[0037] In a fourth aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores at least one instruction, and when the at least one instruction is executed by a processor, the green energy power generation prediction method under the power big data as described above is implemented.

[0038] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0039] This application integrates technologies such as data integration, feature extraction, knowledge graph establishment, graph structure data processing and deep learning model construction, which can analyze green energy systems more comprehensively and accurately, and provide important support and decision-making basis for green energy development and management;

[0040] This application ensures the quality and consistency of data by integrating heterogeneous data from different data sources, cleaning and unifying the format, extracting multi-dimensional features (such as energy features, time features, spatial features, and climate features) from the sorted data, enriching the dimensions and depth of analysis, and establishing a knowledge graph that not only helps understand the complex relationships in the field of green energy, but also provides structured information for advanced analysis and machine learning models;

[0041] This application uses the node vector generated by the Node2Vec algorithm to effectively capture and express the correlation between entities. It combines the graph convolutional network (GCN) and graph attention network (GAT) models, which not only utilizes the structural information of the graph, but also strengthens the model's attention to key nodes through the attention mechanism, which enhances the model's predictive ability and interpretability. The model can predict future energy production based on historical and real-time data, helping energy managers make more scientific decisions. Through feature importance analysis, the main factors affecting energy production can be identified and quantified, providing a basis for formulating strategies.

[0042] The green energy power generation prediction device, electronic device and computer-readable storage medium under power big data provided by the present invention also solve the problems raised in the background technology part. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] The drawings constituting a part of the present application are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0044] Figure 1 This is a flow chart of a method for predicting green energy power generation under power big data according to an embodiment of the present invention;

[0045] Figure 2 This is a structural block diagram of a green energy power generation prediction device under power big data according to an embodiment of the present invention;

[0046] Figure 3 The present invention is a structural block diagram of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0047] The present invention will be described in detail below with reference to the accompanying drawings and in combination with embodiments. It should be noted that the embodiments and features in the embodiments of the present application can be combined with each other without conflict.

[0048] The following detailed description is an exemplary description, which is intended to provide further detailed description of the present invention. Unless otherwise specified, all technical terms used in the present invention have the same meaning as those generally understood by those skilled in the art to which the present application belongs. The terms used in the present invention are only for describing specific embodiments, and are not intended to limit exemplary embodiments according to the present invention.

[0049] Example 1

[0050] like Figure 1 As shown, a green energy power generation prediction method under power big data includes the following steps:

[0051] S1. Obtain heterogeneous data in the power big data platform, including energy production data, meteorological data, and economic indicator data.

[0052] Specifically, heterogeneous data includes energy production data, meteorological data, economic indicators and other data from different data sources.

[0053] Preferably, after acquiring the heterogeneous data, the heterogeneous data is integrated and cleaned to unify the data format and standard.

[0054] As an example, heterogeneous data from different data sources are imported into a unified data lake to ensure centralized storage and management of the data; heterogeneous data are cleaned, including processing missing values, outliers, duplicate values, etc., to ensure the quality and integrity of heterogeneous data; heterogeneous data from different data sources are converted into a unified data format, such as CSV, JSON, Parquet, etc., for subsequent processing and analysis, and heterogeneous data are standardized to ensure that heterogeneous data from different data sources have the same unit and magnitude for comparison and analysis.

[0055] S2. Extracting features from the heterogeneous data, including energy features, time features, space features, and climate features.

[0056] Specifically, use the datetime library in Python to extract time features such as year, season, month, and week; use geographic information system (GIS) tools to convert geographic locations into longitude and latitude or other codes to obtain spatial features; use the meteorological data API to obtain climate features such as temperature, humidity, and precipitation;

[0057] S3. Based on the extracted features, construct a knowledge graph of entities and relationships; wherein the entities include time entities, location entities, climate entities and energy entities, and the relationships include: time-energy relationships, climate-energy relationships and location-energy relationships; convert the entities, relationships and attributes in the knowledge graph into graph structure data, and generate node vector representations based on the graph structure data.

[0058] Specifically, knowledge graphs are used to represent complex relationships in the field of green energy.

[0059] More specifically, the entities include:

[0060] Time entity: including time features such as date and timestamp;

[0061] Location entity: includes geographic location information such as cities;

[0062] Climate entity: including climate data such as temperature, humidity, and precipitation;

[0063] Energy entity: includes energy data such as energy consumption and energy output.

[0064] More specifically, the relationships include:

[0065] Time-Energy Relationship: The relationship between a time entity and an energy entity, described using the energy consumption formula.

[0066] Climate-energy relationship: The relationship between climate entities and energy entities, using the climate impact formula to describe the impact of climate on energy consumption.

[0067] Location-Energy Relationship: The relationship between the location entity and the energy entity, combined with the calculation of the geospatial distance d to describe the impact between locations:

[0068]

[0069] in, , are the latitude and longitude of location 1, , are the latitude and longitude of location 2, is the radius of the Earth.

[0070] Let energy represent energy consumption (unit: kWh), time represent time characteristics, location represent location characteristics, and climate represent climate characteristics. The relationship between energy consumption and time, location, and climate can be expressed as follows:

[0071] energy=f(time,location,climate)

[0072] Climate change has a significant impact on energy demand. Energy demand may increase during colder temperatures, especially in winter, so the following formula is used to estimate temperature-adjusted energy demand:

[0073]

[0074] in, To adjust the energy demand, For basic energy needs, is the adjustment factor, is the standard temperature, is the actual temperature;

[0075] And add time weight attribute to time entity, and calculate according to time weight calculation formula;

[0076]

[0077] Where λ is the decay rate and Δt is the time difference from the current time to the data recording time.

[0078] Specifically, entities, relationships, and attributes in the knowledge graph are converted into graph structure data, and node vector representations are generated based on the graph structure data, including:

[0079] The entities in the knowledge graph are taken as nodes of the graph, and the relationships are taken as edges of the graph. A unique identifier is assigned to each entity and relationship. The attributes of the entity are taken as node features, and the graph is constructed using the graph database Neo4j to obtain graph structure data.

[0080] Based on the Node2Vec algorithm, multiple random walks are performed from each node of the graph, and each random walk generates a node sequence; wherein the random walk is controlled by the return parameter p and the entry and exit parameter q; after the random walk is completed, the random walk sequence and the node sequence corresponding to each random walk are obtained;

[0081] The Word2Vec algorithm is used to convert the node sequence in the random walk sequence into a node vector representation.

[0082] As an example, in the Node2Vec algorithm, the return parameter p and the in-out parameter q are defined to control the random walk strategy. The formula for calculating the transition probability from node v to node t is as follows:

[0083]

[0084] in, P tx represents the probability of transferring from node v to node t, d tx represents the distance from node t to node x, and Z is the normalization factor.

[0085] Specifically, the Word2Vec algorithm is used to convert the node sequence in the random walk sequence into a node vector representation, including:

[0086] The Word2Vec algorithm is used to convert the node sequence in the above walk sequence into a dense vector representation. The Skip-gram model is used to learn the node vector representation. The Skip-gram model aims to learn the vector representation of the node by predicting the neighbor nodes in the node sequence. The generated node sequence is used as input data to train the Skip-gram model to learn the vector representation of the node.

[0087] Furthermore, in the Word2Vec algorithm, the node sequence is used as the input data of the Skip-gram model, and the Skip-gram model learns the vector representation of the node by predicting the neighbor nodes in the node sequence.

[0088] As an example, the Skip-gram model loss function:

[0089]

[0090] Where T is the length of the node sequence, c is the context window size, represents the central node, Represents the context node, Indicates the probability of predicting context node j given the central node;

[0091] In the Skip-gram model, the softmax function is used to calculate the predicted probability:

[0092]

[0093] Among them, u i and v i They represent the input and output vector representations of node i respectively, and V is the total number of nodes.

[0094] S4. Based on the node vector representation, use a pre-trained green energy analysis model to predict future energy production; wherein the green energy analysis model includes a GCN layer and a GAT layer.

[0095] Specifically, the green energy analysis model includes a GCN layer and a GAT layer, wherein:

[0096] The GCN layer is used to utilize the spatial structure information of the graph and propagate features through the adjacency matrix and degree matrix;

[0097] The GAT layer captures the importance of nodes by learning the attention weights between nodes, so that the green energy analysis model pays attention to neighbor nodes whose importance meets the preset requirements. It combines the learned node vector representation as the initial feature and uses the attention mechanism to aggregate the information of neighbor nodes to obtain the trained green energy analysis model.

[0098] More specifically, in this embodiment, the GCN layer is used to propagate spatial structure information, and the adjacency matrix and degree matrix are used for feature aggregation:

[0099]

[0100] in, Indicates that node i is in l Hidden representation of layers; is the Relu activation function, N (i) is the set of neighbor nodes of node i; is the weight matrix of the lth layer; and The degrees of nodes i and j, respectively, are used to plan the characteristics of neighboring nodes; Indicates that node i is in l -1 layer of hidden representation;

[0101] The GAT layer is used to learn the attention weights between nodes, capture the importance between nodes, and focus on neighboring nodes whose importance meets the preset requirements:

[0102]

[0103] in, represents the hidden representation of node i under k attention heads, is the Relu activation function; represents the direct attention weight of nodes i and j, a is the learning parameter in the attention mechanism, is the weight matrix of the k-th attention head in the l-th layer; Indicates that node i is in j Hidden representation of layers; is the weight matrix of the kth attention head; is the hidden representation of node i; is the hidden representation of node j; Represents the concatenation operation of the vector; LeakyReLU is the activation function;

[0104] Combined with the attention mechanism to aggregate the information of neighboring nodes, the model's ability to capture the importance of nodes is improved. The outputs of multiple attention heads are concatenated and residual connections are added to obtain the node representation:

[0105]

[0106] Combining the GCN layer with the GAT layer, we get the final node representation:

[0107] .

[0108] In a preferred embodiment, after the step of using the trained green energy analysis model to predict future energy production, the following steps are further included:

[0109] Using green energy analysis models, identify the main factors affecting energy production;

[0110] Combining the predicted results of energy production and major factors, formulate power generation plans or adjust power generation equipment parameters.

[0111] For example, in this embodiment, a reasonable power generation plan is formulated based on the forecast results of energy consumption and climate impact analysis, including the allocation and time arrangement of power generation. Meteorological data and climate impact formulas are used to predict climate characteristics in the future, so as to adjust the power generation plan to adapt to different climate conditions.

[0112] According to the analysis of the impact of energy consumption, the parameters of power generation equipment are adjusted to improve energy utilization efficiency, and energy consumption is reduced by optimizing the operating parameters of power generation equipment, improving equipment efficiency or introducing new technologies.

[0113] Example 2

[0114] like Figure 2As shown, based on the same inventive concept as the above embodiment, the present invention also provides a green energy power generation prediction device under power big data, which is characterized by comprising:

[0115] Data acquisition module, used to acquire heterogeneous data in the power big data platform, including energy production data, meteorological data and economic indicator data;

[0116] A feature extraction module, used to extract features from the heterogeneous data, including energy features, time features, space features and climate features;

[0117] A graph construction module is used to construct a knowledge graph of entities and relationships based on the extracted features; wherein the entities include time entities, location entities, climate entities and energy entities, and the relationships include: time-energy relationships, climate-energy relationships and location-energy relationships; converting entities, relationships and attributes in the knowledge graph into graph structure data, and generating node vector representations based on the graph structure data;

[0118] The production prediction module is used to predict future energy production based on the node vector representation using a pre-trained green energy analysis model; wherein the green energy analysis model includes a GCN layer and a GAT layer.

[0119] Furthermore, in the graph construction module, entities, relationships, and attributes in the knowledge graph are converted into graph structure data, and node vector representations are generated based on the graph structure data, including the following:

[0120] The entities in the knowledge graph are taken as nodes of the graph, and the relationships are taken as edges of the graph. A unique identifier is assigned to each entity and relationship. The attributes of the entity are taken as node features, and the graph is constructed using the graph database Neo4j to obtain graph structure data.

[0121] Based on the Node2Vec algorithm, multiple random walks are performed from each node of the graph, and each random walk generates a node sequence; wherein the random walk is controlled by the return parameter p and the entry and exit parameter q; after the random walk is completed, the random walk sequence and the node sequence corresponding to each random walk are obtained;

[0122] The Word2Vec algorithm is used to convert the node sequence in the random walk sequence into a node vector representation.

[0123] Furthermore, the Word2Vec algorithm is used to convert the node sequence in the random walk sequence into a node vector representation, including:

[0124] In the Word2Vec algorithm, the node sequence is used as the input data of the Skip-gram model. The Skip-gram model learns the vector representation of the node by predicting the neighbor nodes in the node sequence.

[0125] Example 3

[0126] like Figure 3 As shown, the present invention also provides an electronic device 100 for implementing a method for predicting green energy generation under power big data according to Embodiment 1;

[0127] The electronic device 100 includes a memory 101 , at least one processor 102 , a computer program 103 stored in the memory 101 and executable on the at least one processor 102 , and at least one communication bus 104 .

[0128] The memory 101 can be used to store the computer program 103. The processor 102 implements the steps of a green energy generation prediction method under power big data in Example 1 by running or executing the computer program stored in the memory 101 and calling the data stored in the memory 101.

[0129] The memory 101 may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function (such as a sound playback function, an image playback function, etc.), etc.; the data storage area may store data (such as audio data) created according to the use of the electronic device 100, etc. In addition, the memory 101 may include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), at least one disk storage device, a flash memory device, or other non-volatile solid-state storage devices.

[0130] At least one processor 102 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The processor 102 may be a microprocessor or any conventional processor, etc. The processor 102 is the control center of the electronic device 100, and uses various interfaces and lines to connect various parts of the entire electronic device 100.

[0131] The memory 101 in the electronic device 100 stores a plurality of instructions to implement a method for predicting green energy generation under power big data, and the processor 102 can execute the plurality of instructions to implement:

[0132] Acquire heterogeneous data in the power big data platform, including energy production data, meteorological data, and economic indicator data;

[0133] Extracting features from the heterogeneous data, including energy features, time features, spatial features, and climate features;

[0134] Based on the extracted features, a knowledge graph of entities and relationships is constructed; wherein the entities include time entities, location entities, climate entities and energy entities, and the relationships include: time-energy relationships, climate-energy relationships and location-energy relationships; entities, relationships and attributes in the knowledge graph are converted into graph structure data, and node vector representations are generated based on the graph structure data;

[0135] Based on the node vector representation, a pre-trained green energy analysis model is used to predict future energy production; wherein the green energy analysis model includes a GCN layer and a GAT layer.

[0136] Example 4

[0137] If the module / unit integrated in the electronic device 100 is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present invention implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. Computer-readable media may include: any entity or device capable of carrying computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory and read-only memory (ROM, Read-Only Memory).

[0138] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0139] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0140] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0141] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0142] In the description of this specification, the description with reference to the terms "one embodiment", "example", "specific example", etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0143] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A green energy generation prediction method based on power big data, characterized in that: The following steps are involved: Acquire heterogeneous data in the power big data platform, including energy production data and meteorological data; Extracting features from the heterogeneous data, including energy features, time features, spatial features, and climate features; Based on the extracted features, a knowledge graph of entities and relationships is constructed; wherein the entities include time entities, location entities, climate entities and energy entities, and the relationships include: time-energy relationships, climate-energy relationships and location-energy relationships; entities, relationships and attributes in the knowledge graph are converted into graph structure data, and node vector representations are generated based on the graph structure data; Based on the node vector representation, a pre-trained green energy analysis model is used to predict future energy production; wherein the green energy analysis model includes a GCN layer and a GAT layer; Converting entities, relationships, and attributes in the knowledge graph into graph structure data and generating node vector representations based on the graph structure data includes the following steps: The entities in the knowledge graph are taken as nodes of the graph, and the relationships are taken as edges of the graph. A unique identifier is assigned to each entity and relationship. The attributes of the entity are taken as node features, and the graph is constructed using the graph database Neo4j to obtain graph structure data. Based on the Node2Vec algorithm, multiple random walks are performed from each node of the graph, and each random walk generates a node sequence; wherein the random walk is controlled by the return parameter p and the entry and exit parameter q; after the random walk is completed, the random walk sequence and the node sequence corresponding to each random walk are obtained; The Word2Vec algorithm is used to convert the node sequence in the random walk sequence into a node vector representation.

2. The green energy generation prediction method under power big data according to claim 1 is characterized in that: Using the Word2Vec algorithm, the node sequence in the random walk sequence is converted into a node vector representation, including: In the Word2Vec algorithm, the node sequence is used as the input data of the Skip-gram model. The Skip-gram model learns the vector representation of the node by predicting the neighbor nodes in the node sequence.

3. The green energy generation prediction method under power big data according to claim 2 is characterized in that: The green energy analysis model includes a GCN layer and a GAT layer, wherein: The GCN layer is used to utilize the spatial structure information of the graph and propagate features through the adjacency matrix and degree matrix; The GAT layer captures the importance of nodes by learning the attention weights between nodes, so that the green energy analysis model pays attention to neighbor nodes whose importance meets the preset requirements. It combines the learned node vector representation as the initial feature and uses the attention mechanism to aggregate the information of neighbor nodes to obtain the trained green energy analysis model.

4. The green energy generation prediction method under power big data according to claim 1 is characterized in that: After the step of using the trained green energy analysis model to predict future energy production, it also includes: Using green energy analysis models, identify the main factors affecting energy production; Combining the predicted results of energy production and major factors, formulate power generation plans or adjust power generation equipment parameters.

5. A green energy generation prediction device under power big data, characterized in that: include: Data acquisition module, used to acquire heterogeneous data in the power big data platform, including energy production data and meteorological data; A feature extraction module, used to extract features from the heterogeneous data, including energy features, time features, space features and climate features; A graph construction module is used to construct a knowledge graph of entities and relationships based on the extracted features; wherein the entities include time entities, location entities, climate entities and energy entities, and the relationships include: time-energy relationships, climate-energy relationships and location-energy relationships; converting entities, relationships and attributes in the knowledge graph into graph structure data, and generating node vector representations based on the graph structure data; A yield prediction module, used to predict future energy yields based on the node vector representation using a pre-trained green energy analysis model; wherein the green energy analysis model includes a GCN layer and a GAT layer; In the graph construction module, entities, relationships, and attributes in the knowledge graph are converted into graph structure data, and node vector representations are generated based on the graph structure data, including the following: The entities in the knowledge graph are taken as nodes of the graph, and the relationships are taken as edges of the graph. A unique identifier is assigned to each entity and relationship. The attributes of the entity are taken as node features, and the graph is constructed using the graph database Neo4j to obtain graph structure data. Based on the Node2Vec algorithm, multiple random walks are performed from each node of the graph, and each random walk generates a node sequence; wherein the random walk is controlled by the return parameter p and the entry and exit parameter q; after the random walk is completed, the random walk sequence and the node sequence corresponding to each random walk are obtained; The Word2Vec algorithm is used to convert the node sequence in the random walk sequence into a node vector representation.

6. The green energy generation prediction device under power big data according to claim 5 is characterized in that: Using the Word2Vec algorithm, the node sequence in the random walk sequence is converted into a node vector representation, including: In the Word2Vec algorithm, the node sequence is used as the input data of the Skip-gram model. The Skip-gram model learns the vector representation of the node by predicting the neighbor nodes in the node sequence.

7. An electronic device, characterized in that: It comprises a processor and a memory, wherein the processor is used to execute a computer program stored in the memory to implement the green energy power generation prediction method under power big data as described in any one of claims 1 to 4.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores at least one instruction, and when the at least one instruction is executed by the processor, the green energy power generation prediction method under power big data as described in any one of claims 1 to 4 is implemented.

Images