Verification method, system and device for green certificate issuing multi-source data and medium

Abstract

The invention relates to the technical field of energy data management, in particular to a verification method, system and device for green certificate issuing multi-source data and a medium. The method comprises the following steps of: firstly, acquiring green certificate issuing multi-source data according to a constructed multi-source heterogeneous data interface layer; establishing a data warehouse according to distributed storage, and calling a project information mapping algorithm to associate cross-system green certificate issuing multi-source data; and finally, mapping the relation graph to obtain geographic grid codes, constructing an inter-provincial feature matrix, and checking inter-provincial green certificate issuing multi-source data. Through deep integration of a layered architecture and a tool chain, a three-layer multi-source heterogeneous data interface is constructed, full-amount collection of basic data of various types of renewable energy power generation projects is realized, and the integrity, accuracy and verification efficiency of green certificate verification data are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the field of energy data management technology, and in particular to a method, system, device and medium for verifying multi-source data for issuing green certificates. Background Art

[0002] Currently, the renewable energy power market is rapidly expanding, and the scope of green certificate issuance is rapidly expanding. The amount of data generated throughout the entire issuance process is rapidly increasing, and the requirements for issuance verification are becoming increasingly stringent. Existing data collection methods and verification technologies are unable to support the collection and integrity verification of massive amounts of data. There is an urgent need to develop intelligent and efficient multi-source data collection and verification technologies for green certificate issuance, promote the efficient circulation of green certificate data, and promote full coverage of green certificate issuance.

[0003] Currently, the Green Certificate trading system only aggregates data from the marketing system and cannot capture the full range of renewable energy project data under the full scope of Green Certificate issuance. Existing data aggregation technology cannot support the system's multi-source data aggregation, making it difficult to map and match project information. Currently, renewable energy entity information and project archive information are collected from multiple systems. In practice, data entry, updating, and review processes are imperfect, resulting in inconsistent data quality and numerous issues. While national review rules have addressed most data quality issues, verification of complex business logic remains insufficient. Summary of the Invention

[0004] In response to the problem that the data entry, update and review processes of the existing green certificate trading system are imperfect in actual operation, resulting in uneven data quality, the present invention proposes a verification method, system, equipment and medium for multi-source data of green certificate issuance; the method first collects multi-source data of green certificate issuance based on the constructed multi-source heterogeneous data interface layer; then establishes a data warehouse based on distributed storage, and calls the project information mapping algorithm to associate multi-source data of green certificate issuance across systems; finally, the mapping relationship map obtains geographic grid codes, constructs an inter-provincial feature matrix, and verifies multi-source data of green certificate issuance between provinces; significantly improving the integrity, accuracy and verification efficiency of green certificate issuance data.

[0005] The specific implementation contents of the present invention are as follows:

[0006] A method for verifying multi-source data for issuing green certificates, comprising the following steps:

[0007] Step S1: Collect multi-source data for green certificate issuance based on the constructed multi-source heterogeneous data interface layer;

[0008] Step S2: Establish a data warehouse based on distributed storage, and call the project information mapping algorithm to associate the multi-source data of green certificate issuance across systems to generate a mapping relationship map;

[0009] Step S3: Obtain geographic grid codes based on the mapping relationship map, construct an inter-provincial feature matrix, and verify the multi-source data on inter-provincial green certificate issuance.

[0010] In order to better implement the present invention, further, step S1 specifically includes the following steps:

[0011] Step S11: Build a three-layer multi-source heterogeneous data interface based on the configured API gateway and ETL tool chain;

[0012] Step S12: Collect renewable energy project data from the marketing system, incremental distribution network management system and local power grid data platform according to the constructed three-layer multi-source heterogeneous data interface to obtain multi-source data for green certificate issuance.

[0013] In order to better implement the present invention, further, step S11 specifically includes the following steps:

[0014] Step S111: Configure the API gateway according to the upstream defined by the Kong gateway and the constructed routing;

[0015] Step S112: Based on the configured API gateway, call the beam.Pipeline() function to obtain a token and build an ETL pipeline;

[0016] Step S113: Build a GraphQL service layer query interface based on the Customer type and MetricType enumeration, set up a Redis Cluster cache to store hot data, and set up a Caffeine local cache to store cold data;

[0017] Step S114: Build a three-layer multi-source heterogeneous data interface based on the configured API gateway, ETL pipeline, and service layer query interface.

[0018] In order to better implement the present invention, further, the specific operations of step S12 are:

[0019] The collection strategy is dynamically adjusted according to the constructed credibility assessment model, and renewable energy project data is collected from the marketing system, incremental distribution network management system and local power grid data platform based on the constructed three-layer multi-source heterogeneous data interface to obtain multi-source data for green certificate issuance.

[0020] In order to better implement the present invention, further, step S2 specifically includes the following steps:

[0021] Step S21: Filter the collected multi-source data on green certificate issuance according to the configured standardized rule base;

[0022] Step S22: Based on distributed storage, a data warehouse including a time series database and a graph database is constructed to store the filtered multi-source data on green certificate issuance;

[0023] Step S23: Obtain project geographic data from multi-source data on green certificate issuance, and use the multi-attention mechanism to calculate similarity and generate a mapping relationship map;

[0024] Step S24: Correlate cross-system renewable energy project data according to the mapping relationship graph.

[0025] In order to better implement the present invention, further, step S23 specifically includes the following steps:

[0026] Step S231: Acquire project geographic data from multi-source data for green certificate issuance and convert it into a unified coordinate system, and use NLP technology to parse the address hierarchy of the unified project geographic data;

[0027] Step S232: Based on the set spatial threshold, call the GeoHash() function to generate a spatial neighbor candidate set;

[0028] Step S233: Call the Haversine() function to calculate the spherical distance based on the spatial neighbor candidate set to obtain the geometric similarity;

[0029] Step S234: calling the SentenceTransformer model according to the address level to calculate the address semantic similarity;

[0030] Step S235: Call the torch.nn.Module module to build a graph neural network, and call the forward() function to learn the topological structure of the spatial neighboring candidate set in the graph to obtain the graph embedding similarity;

[0031] Step S236: Based on the set learning weight parameters, geometric similarity, address semantic similarity, and graph embedding similarity, the attention mechanism is called to obtain multimodal similarity;

[0032] Step S237: Call the XGBClassifier() function to train multimodal similarity and generate a mapping relationship graph based on the set cross-graph relationship constraints.

[0033] In order to better implement the present invention, further, step S3 specifically includes the following steps:

[0034] Step S31: Obtaining the geographic grid code according to the mapping relationship map, and combining it with the obtained power generation entity ID to generate a composite primary key;

[0035] Step S32: constructing an inter-provincial characteristic matrix based on the set inter-provincial output correlation coefficient, geographical attenuation factor, and constructed composite primary key;

[0036] Step S33: Based on the inter-provincial feature matrix and the set contamination parameters, the Isolation Forest() function is called to determine whether the multi-source data of inter-provincial green certificate issuance is abnormal. If abnormal, an alarm is generated, isolation is carried out, and blockchain evidence storage is triggered; otherwise, step S34 is executed;

[0037] Step S34: Call the GTA network to calculate the edge attention weight of the abnormal transaction path. If the edge attention weight is less than the set weight threshold, the multi-source data of the inter-provincial green certificate issuance is judged to be suspicious and transmitted to the manual review queue, and the related transactions are frozen at the same time; otherwise, execute step S35;

[0038] Step S35: Call the LSTM-Attention network to predict the power generation in the next period. If the power generation is less than the deviation threshold, log it and generate a data quality report.

[0039] Based on the above-mentioned verification method for multi-source data of green certificate issuance, in order to better implement the present invention, a verification system for multi-source data of green certificate issuance is further proposed, which is used to execute the above-mentioned verification method for multi-source data of green certificate issuance; comprising a multi-source data acquisition unit, a mapping unit, and a verification unit;

[0040] The multi-source data acquisition unit is used to collect multi-source data for green certificate issuance based on the constructed multi-source heterogeneous data interface layer;

[0041] The mapping unit is used to establish a data warehouse based on distributed storage and call the project information mapping algorithm to associate the multi-source data of green certificate issuance across systems;

[0042] The verification unit is used to verify the associated inter-provincial green certificate issuance multi-source data based on the set intelligent verification engine fusion rules and the constructed machine learning model.

[0043] Based on the above-mentioned verification method for multi-source data for issuing green certificates, in order to better realize the present invention, an electronic device is further proposed, including a memory and a processor; a computer program is stored on the memory; when the computer program is executed on the processor, the above-mentioned verification method for multi-source data for issuing green certificates is implemented.

[0044] Based on the above-mentioned verification method for multi-source data for green certificate issuance, in order to better realize the present invention, a computer-readable storage medium is further proposed, on which computer instructions are stored; when the computer instructions are executed on the above-mentioned electronic device, the above-mentioned verification method for multi-source data for green certificate issuance is implemented.

[0045] The present invention has the following beneficial effects:

[0046] (1) The present invention builds a three-layer multi-source heterogeneous data interface through deep integration of layered architecture and tool chain, realizes the full collection of basic data of various types of renewable energy power generation projects across the country, and processes hundreds of millions of heterogeneous data requests per day.

[0047] (2) The present invention realizes the association of multi-source data for green certificate issuance across systems by calling the project information mapping algorithm, laying a data management foundation for the issuance of green certificates.

[0048] (3) The present invention constructs an inter-provincial feature matrix, verifies the multi-source data of inter-provincial green certificate issuance, realizes the automatic data verification of inter-provincial business logic, improves the data quality, and ensures the accuracy and consistency of the data. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 A flow chart of the method for verifying multi-source data for issuing green certificates provided by the present invention.

[0050] Figure 2 Schematic diagram of the low-dimensional embedding process of the graph neural network provided by the present invention. DETAILED DESCRIPTION

[0051] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. It should be understood that the described embodiments are only part of the embodiments of the present invention, not all of the embodiments, and therefore should not be regarded as limiting the scope of protection. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technical personnel in this field without making creative work are within the scope of protection of the present invention.

[0052] In the description of the present invention, it should be noted that, unless otherwise expressly specified or limited, the terms "disposed," "connected," and "connected" should be understood in a broad sense. For example, they may refer to fixed connections, detachable connections, or integral connections; mechanical connections, electrical connections; direct connections, indirect connections through an intermediate medium, or internal connections between two components. Those skilled in the art will understand the specific meanings of the above terms in the present invention based on the specific circumstances.

[0053] Example 1:

[0054] This embodiment proposes a verification method for multi-source data for green certificate issuance, such as Figure 1 As shown, the specific steps include:

[0055] Step S1: Collect multi-source data for green certificate issuance based on the constructed multi-source heterogeneous data interface layer.

[0056] The step S1 specifically includes the following steps:

[0057] Step S11: Build a three-layer multi-source heterogeneous data interface based on the configured API gateway and ETL tool chain;

[0058] The step S11 specifically includes the following steps:

[0059] Step S111: Configure the API gateway according to the upstream defined by the Kong gateway and the constructed routing;

[0060] Step S112: Based on the configured API gateway, call the beam.Pipeline() function to obtain a token and build an ETL pipeline;

[0061] Step S113: Build a GraphQL service layer query interface based on the Customer type and MetricType enumeration, set up a Redis Cluster cache to store hot data, and set up a Caffeine local cache to store cold data;

[0062] Step S114: Build a three-layer multi-source heterogeneous data interface based on the configured API gateway, ETL pipeline, and service layer query interface.

[0063] Step S12: Collect renewable energy project data from the marketing system, incremental distribution network management system and local power grid data platform according to the constructed three-layer multi-source heterogeneous data interface to obtain multi-source data for green certificate issuance.

[0064] The specific operations of step S12 are:

[0065] The collection strategy is dynamically adjusted according to the constructed credibility assessment model, and renewable energy project data is collected from the marketing system, incremental distribution network management system and local power grid data platform based on the constructed three-layer multi-source heterogeneous data interface to obtain multi-source data for green certificate issuance.

[0066] Step S2: Establish a data warehouse based on distributed storage, and call the project information mapping algorithm to associate the multi-source data of green certificate issuance across systems to generate a mapping relationship map.

[0067] The step S2 specifically includes the following steps:

[0068] Step S21: Filter the collected multi-source data on green certificate issuance according to the configured standardized rule base;

[0069] Step S22: Based on distributed storage, a data warehouse including a time series database and a graph database is constructed to store the filtered multi-source data on green certificate issuance;

[0070] Step S23: Obtain project geographic data from multi-source data on green certificate issuance, and use the multi-attention mechanism to calculate similarity and generate a mapping relationship map;

[0071] The step S23 specifically includes the following steps:

[0072] Step S231: Acquire project geographic data from multi-source data for green certificate issuance and convert it into a unified coordinate system, and use NLP technology to parse the address hierarchy of the unified project geographic data;

[0073] Step S232: Based on the set spatial threshold, call the GeoHash() function to generate a spatial neighbor candidate set;

[0074] Step S233: Call the Haversine() function to calculate the spherical distance based on the spatial neighbor candidate set to obtain the geometric similarity;

[0075] Step S234: calling the SentenceTransformer model according to the address level to calculate the address semantic similarity;

[0076] Step S235: Call the torch.nn.Module module to build a graph neural network, and call the forward() function to learn the topological structure of the spatial neighboring candidate set in the graph to obtain the graph embedding similarity;

[0077] Step S236: Based on the set learning weight parameters, geometric similarity, address semantic similarity, and graph embedding similarity, the attention mechanism is called to obtain multimodal similarity;

[0078] Step S237: Call the XGBClassifier() function to train multimodal similarity and generate a mapping relationship graph based on the set cross-graph relationship constraints.

[0079] Step S24: Correlate cross-system renewable energy project data according to the mapping relationship graph.

[0080] Step S3: Obtain geographic grid codes based on the mapping relationship map, construct an inter-provincial feature matrix, and verify the multi-source data on inter-provincial green certificate issuance.

[0081] The step S3 specifically includes the following steps:

[0082] Step S31: Obtaining the geographic grid code according to the mapping relationship map, and combining it with the obtained power generation entity ID to generate a composite primary key;

[0083] Step S32: constructing an inter-provincial characteristic matrix based on the set inter-provincial output correlation coefficient, geographical attenuation factor, and constructed composite primary key;

[0084] Step S33: Based on the inter-provincial feature matrix and the set contamination parameters, the Isolation Forest() function is called to determine whether the multi-source data of inter-provincial green certificate issuance is abnormal. If abnormal, an alarm is generated, isolation is carried out, and blockchain evidence storage is triggered; otherwise, step S34 is executed;

[0085] Step S34: Call the GTA network to calculate the edge attention weight of the abnormal transaction path. If the edge attention weight is less than the set weight threshold, the multi-source data of the inter-provincial green certificate issuance is judged to be suspicious and transmitted to the manual review queue, and the related transactions are frozen at the same time; otherwise, execute step S35;

[0086] Step S35: Call the LSTM-Attention network to predict the power generation in the next period. If the power generation is less than the deviation threshold, log it and generate a data quality report.

[0087] Working principle: This embodiment first collects multi-source data on green certificate issuance based on the constructed multi-source heterogeneous data interface layer; then establishes a data warehouse based on distributed storage, and calls the project information mapping algorithm to associate the multi-source data on green certificate issuance across systems; finally, the mapping relationship map obtains the geographic grid code, constructs the inter-provincial feature matrix, and verifies the multi-source data on green certificate issuance between provinces; significantly improves the integrity, accuracy and verification efficiency of green certificate issuance data.

[0088] Example 2:

[0089] This embodiment is based on the above embodiment 1. Figure 2 As shown, a specific embodiment is described in detail, which specifically includes the following steps.

[0090] Step S1: Collect multi-source data for green certificate issuance based on the constructed multi-source heterogeneous data interface layer.

[0091] Step S11: Build a three-layer multi-source heterogeneous data interface based on the configured API gateway and ETL tool chain;

[0092] Step S111: Configure the API gateway according to the upstream defined by the Kong gateway and the constructed routing;

[0093] This example first defines an upstream in Kong, then creates a route to match requests for the path. After configuration, the request is sent to the Kong proxy and forwarded to the backend, implementing multi-backend routing configuration. At the same time, the gRPC-JSON converter and XML-to-JSON conversion middleware are configured as protocol conversion plug-ins.

[0094] Step S112: Based on the configured API gateway, call the beam.Pipeline() function to build the ETL pipeline.

[0095] In this embodiment, based on the configured API, a Lambda function is called on the backend as the integration type, and the beam.Pipeline() function is called based on the Lambda function to build an ETL pipeline.

[0096] Step S113: Build a GraphQL service layer query interface based on the Customer type and MetricType enumeration, set up a Redis Cluster cache to store hot data, and set up a Caffeine local cache to store cold data;

[0097] Step S114: Build a three-layer multi-source heterogeneous data interface based on the configured API gateway, ETL pipeline, and service layer query interface.

[0098] This embodiment achieves daily processing of hundreds of millions of heterogeneous data requests through deep integration of layered architecture and tool chain, while ensuring that the data service SLA reaches above 99.95%, with a single API endpoint throughput of ≥5000TPS and an end-to-end latency P99 <300ms.

[0099] Step S12: Collect renewable energy project data from the marketing system, incremental distribution network management system and local power grid data platform according to the constructed three-layer multi-source heterogeneous data interface to obtain multi-source data for green certificate issuance.

[0100] The specific operations of step S12 are:

[0101] The collection strategy is dynamically adjusted according to the constructed credibility assessment model, and renewable energy project data is collected from the marketing system, incremental distribution network management system and local power grid data platform based on the constructed three-layer multi-source heterogeneous data interface to obtain multi-source data for green certificate issuance.

[0102] Step S2: Establish a data warehouse based on distributed storage, and call the project information mapping algorithm to associate the multi-source data of green certificate issuance across systems to generate a mapping relationship map.

[0103] The step S2 specifically includes the following steps:

[0104] Step S21: Filter the collected multi-source data on green certificate issuance according to the configured standardized rule base;

[0105] Step S22: Based on distributed storage, a data warehouse including a time series database and a graph database is constructed to store the filtered multi-source data on green certificate issuance;

[0106] Step S23: Obtain project geographic data from multi-source data on green certificate issuance, and use the multi-attention mechanism to calculate similarity and generate a mapping relationship map;

[0107] The step S23 specifically includes the following steps:

[0108] Step S231: Acquire project geographic data from multi-source data for green certificate issuance and convert it into a unified coordinate system, and use NLP technology to parse the address hierarchy of the unified project geographic data;

[0109] In this embodiment, the project geographic data is first converted to a unified coordinate system such as the WGS84 coordinate system. If an encrypted coordinate system exists, the encrypted coordinate system needs to be reverse-corrected using GCJ-02. Then, the address is standardized, i.e., the address hierarchy is parsed using NLP technology. For example, if the original address is "Province A, City B, District C, Street D, No. E", the parsed address hierarchy is "Province A", "City B", "District C", "Street D", and "No. E".

[0110] Step S232: Based on the set spatial threshold, call the GeoHash() function to generate a spatial neighbor candidate set;

[0111] In this embodiment, a spatial threshold is set, for example, the spatial threshold is set to 1, and the unit is km; then, based on the longitude and latitude coordinates obtained from the geographic data and the set grid number m, the GeoHash() function is called to quickly filter the spatial neighboring candidate set F{f1, f2, f3, ..., f n}, generate m-bit geohash grid;

[0112] Step S233: Call the Haversine() function to calculate the spherical distance based on the spatial neighbor candidate set to obtain the geometric similarity;

[0113] From the spatial neighboring candidate set F{f1, f2, f3, ..., f n} to obtain the first geographic data f1(lat1,lon1) and the second geographic data f 21 (lat2,lon2); call Haversine() function to calculate the spherical distance d;

[0114] dlon=lon2-lon1;

[0115] dlat=lat2-lat1;

[0116] d=2*racsin(√(sin 2 (dlat / 2)+cos(lat1)*cos(lat2)*sin 2 (dlon / 2)));

[0117] Where r is the radius of the Earth, dlon is the difference in longitude, and dlat is the difference in latitude.

[0118] Step S234: calling the SentenceTransformer model according to the address level to calculate the address semantic similarity;

[0119] In this embodiment, the parsed address levels of "Province A", "City B", "District C", "Street D", and "No. E" are input into the SentenceTransformer model, and each address level is encoded separately. The cosine similarity between the address levels is calculated, and the final semantic similarity is calculated based on the set weights. By calculating the semantic similarity of the addresses, abbreviations, aliases, and addresses with misplaced sequences can be identified, such as "Haidian District, Beijing" and "Haidian District, Beijing", which are identified as the same address based on the semantic similarity calculation.

[0120] Step S235: Call the torch.nn.Module module to build a graph neural network, and call the forward() function to learn the topological structure of the spatial neighboring candidate set in the graph to obtain the graph embedding similarity;

[0121] This embodiment directly introduces the torch.nn.Module module in the forward() function, uses the address level as a node, constructs an edge index based on the edges between the address levels, generates a low-dimensional graph embedding through the first graph convolution layer conv1 and the second graph convolution layer conv2, then calls the forward() function to aggregate the low-dimensional graph embedding to generate an embedding vector G; finally, the graph embedding similarity is calculated based on the normalized embedding vector.

[0122]

[0123] Gnorm is the embedding vector after L2 normalization;

[0124]

[0125] Among them, Gnorm1 and Gnorm2 are two normalized embedding vectors, G Similarity Embedding similarity for graphs.

[0126] Step S236: Based on the set learning weight parameters, geometric similarity, address semantic similarity, and graph embedding similarity, the attention mechanism is called to obtain multimodal similarity;

[0127] similarity = α*(geometric similarity) + β*(semantic similarity) + γ*(graph embedding similarity);

[0128] Among them, α, β, and γ are learning weight parameters, and α+β+γ=1.

[0129] Step S237: Call the XGBClassifier() function to train multimodal similarity and generate a mapping relationship graph based on the set cross-graph relationship constraints.

[0130] Call the XGBClassifier() function to train multimodal similarity, use the predicted probability output by the model as the similarity score, calculate the predicted probability of cross-graph multimodal similarity based on the trained XGBoost model, and optimize according to the set threshold and set constraints to obtain the entity pair with the highest predicted probability. The constraint set in this embodiment is that one entity matches at most one entity in another graph.

[0131] Step S24: Correlate cross-system renewable energy project data according to the mapping relationship graph.

[0132] This embodiment uses a multimodal entity alignment framework to fuse coordinate geometry features, text semantic features, and graph structure features; the fusion of multi-source features achieves a matching accuracy of over 95%.

[0133] Step S3: Obtain geographic grid codes based on the mapping relationship map, construct an inter-provincial feature matrix, and verify the multi-source data on inter-provincial green certificate issuance.

[0134] The step S3 specifically includes the following steps:

[0135] Step S31: Obtain the geographic grid code according to the mapping relationship map, combine it with the obtained power generation entity ID, and generate a composite primary key; and call Tukey's fences method (IQR = 1.5) to remove outliers in numerical fields such as power generation and transaction volume;

[0136] Step S32: constructing an inter-provincial characteristic matrix based on the set inter-provincial output correlation coefficient, geographical attenuation factor, and constructed composite primary key;

[0137] Step S33: Based on the inter-provincial feature matrix and the set contamination parameter, the Isolation Forest() function is called to determine whether the multi-source data of inter-provincial green certificate issuance is abnormal. If contamination = 0.05, an alarm is generated for the abnormal point, isolation is carried out, and blockchain evidence storage is triggered; otherwise, step S34 is executed;

[0138] Step S34: Call the GTA network to calculate the edge attention weight of the abnormal transaction path. If the edge attention weight is less than the set weight threshold, that is, the edge attention weight is less than 0.1, the multi-source data of the inter-provincial green certificate issuance is judged to be suspicious and transmitted to the manual review queue, and the related transactions are frozen at the same time; otherwise, execute step S35;

[0139] Step S35: Call the LSTM-Attention network to predict the power generation in the next period. If the power generation is less than the deviation threshold, a log is recorded and a data quality report is generated. The following table shows the hierarchical processing operations corresponding to abnormal data.

[0140] Table 1 Abnormal operation corresponding graded processing table

[0141] grade operate Level 1 serious error Directly isolate and trigger blockchain evidence Level 2 suspected anomaly Enter the manual review queue and freeze related transactions at the same time Level 3 slight deviation Record logs and generate data quality reports

[0142] Through the above embodiment, based on the measured data of a pilot project, real-time verification of more than 100 million inter-provincial green certificate data can be achieved on a daily basis, with an accuracy rate of up to 99.2%, and the workload of manual review can be reduced by 73%.

[0143] The rest of this embodiment is the same as that of the above-mentioned embodiment 1, and therefore will not be described in detail.

[0144] Example 3:

[0145] This embodiment, based on any one of the above embodiments 1-2, further provides a verification system for multi-source data for green certificate issuance, which is used to execute the above-mentioned verification method for multi-source data for green certificate issuance; comprising a multi-source data acquisition unit, a mapping unit, and a verification unit;

[0146] The multi-source data acquisition unit is used to collect multi-source data for green certificate issuance based on the constructed multi-source heterogeneous data interface layer;

[0147] The mapping unit is used to establish a data warehouse based on distributed storage and call the project information mapping algorithm to associate the multi-source data of green certificate issuance across systems;

[0148] The verification unit is used to verify the associated inter-provincial green certificate issuance multi-source data based on the set intelligent verification engine fusion rules and the constructed machine learning model.

[0149] This embodiment also proposes an electronic device, including a memory and a processor; a computer program is stored on the memory; when the computer program is executed on the processor, the above-mentioned verification method for multi-source data for green certificate issuance is implemented.

[0150] This embodiment also proposes a computer-readable storage medium, on which computer instructions are stored; when the computer instructions are executed on the above-mentioned electronic device, the above-mentioned method for verifying multi-source data for issuing green certificates is implemented.

[0151] The rest of this embodiment is the same as any of the above-mentioned embodiments 1 and 2, and thus will not be described in detail.

[0152] The processor involved in the embodiments of the present application may be a chip. For example, it may be a field programmable gate array (FPGA), an application specific integrated circuit (ASIC), a system on chip (SoC), a central processor unit (CPU), a network processor (NP), a digital signal processor (DSP), a microcontroller unit (MCU), a programmable logic device (PLD), or other integrated chips.

[0153] The memory involved in the embodiments of the present application may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memories. Among them, the non-volatile memory may be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory may be a random access memory (RAM), which is used as an external cache. By way of example and not limitation, many forms of RAM are available, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate synchronous DRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), and direct RAM bus RAM (DR RAM). It should be noted that the memory of the systems and methods described herein is intended to include, but is not limited to, these and any other suitable types of memory.

[0154] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0155] Those skilled in the art will appreciate that the modules and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0156] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and modules described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0157] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of the modules is merely a logical function division. In actual implementation, there may be other division methods, such as multiple modules or components can be combined or integrated into another device, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or modules, which can be electrical, mechanical or other forms.

[0158] The modules described as separate components may or may not be physically separate, and the components shown as modules may or may not be physical modules, that is, they may be located on a single device or distributed across multiple devices. Some or all of the modules may be selected to achieve the purpose of this embodiment based on actual needs.

[0159] In addition, the functional modules in the various embodiments of the present application may be integrated into one device, or each module may exist physically separately, or two or more modules may be integrated into one device.

[0160] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented using a software program, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When loading and executing computer program instructions on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from a website, computer, server or data center by wired (such as coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (such as infrared, wireless, microwave, etc.) mode to another website, computer, server or data center. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more media that can be integrated. The available medium may be a magnetic medium (eg, a floppy disk, a hard disk, a magnetic tape), an optical medium (eg, a DVD), or a semiconductor medium (eg, a solid state disk (SSD)).

[0161] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

Claims

1. A method for verifying multi-source data for issuing green certificates, characterized in that: The specific steps include: Step S1: Collect multi-source data for green certificate issuance based on the constructed multi-source heterogeneous data interface layer; Step S2: Establish a data warehouse based on distributed storage, and call the project information mapping algorithm to associate the multi-source data of green certificate issuance across systems to generate a mapping relationship map; Step S3: Obtain geographic grid codes based on the mapping relationship map, construct an inter-provincial feature matrix, and verify the multi-source data on inter-provincial green certificate issuance.

2. A method for verifying multi-source data for green certificate issuance according to claim 1, characterized in that: The step S1 specifically includes the following steps: Step S11: Build a three-layer multi-source heterogeneous data interface based on the configured API gateway and ETL tool chain; Step S12: Collect renewable energy project data from the marketing system, incremental distribution network management system and local power grid data platform according to the constructed three-layer multi-source heterogeneous data interface to obtain multi-source data for green certificate issuance.

3. A method for verifying multi-source data for green certificate issuance according to claim 2, characterized in that: The step S11 specifically includes the following steps: Step S111: Configure the API gateway according to the upstream defined by the Kong gateway and the constructed routing; Step S112: Based on the configured API gateway, call the beam.Pipeline() function to obtain a token and build an ETL pipeline; Step S113: Build a GraphQL service layer query interface based on the Customer type and MetricType enumeration, set up a Redis Cluster cache to store hot data, and set up a Caffeine local cache to store cold data; Step S114: Build a three-layer multi-source heterogeneous data interface based on the configured API gateway, ETL pipeline, and service layer query interface.

4. A method for verifying multi-source data for green certificate issuance according to claim 2, characterized in that: The specific operations of step S12 are: The collection strategy is dynamically adjusted according to the constructed credibility assessment model, and renewable energy project data is collected from the marketing system, incremental distribution network management system and local power grid data platform based on the constructed three-layer multi-source heterogeneous data interface to obtain multi-source data for green certificate issuance.

5. A method for verifying multi-source data for green certificate issuance according to claim 2, characterized in that: The step S2 specifically includes the following steps: Step S21: Filter the collected multi-source data on green certificate issuance according to the configured standardized rule base; Step S22: Based on distributed storage, a data warehouse including a time series database and a graph database is constructed to store the filtered multi-source data on green certificate issuance; Step S23: Obtain project geographic data from multi-source data on green certificate issuance, and use the multi-attention mechanism to calculate similarity and generate a mapping relationship map; Step S24: Correlate cross-system renewable energy project data according to the mapping relationship graph.

6. A method for verifying multi-source data for green certificate issuance according to claim 5, characterized in that: The step S23 specifically includes the following steps: Step S231: Acquire project geographic data from multi-source data for green certificate issuance and convert it into a unified coordinate system, and use NLP technology to parse the address hierarchy of the unified project geographic data; Step S232: Based on the set spatial threshold, call the GeoHash() function to generate a spatial neighbor candidate set; Step S233: Call the Haversine() function to calculate the spherical distance based on the spatial neighbor candidate set to obtain the geometric similarity; Step S234: Call the SentenceTransformer model according to the address level to calculate the address semantic similarity; Step S235: Call the torch.nn.Module module to build a graph neural network, and call the forward() function to learn the topological structure of the spatial neighboring candidate set in the graph to obtain the graph embedding similarity; Step S236: Based on the set learning weight parameters, geometric similarity, address semantic similarity, and graph embedding similarity, the attention mechanism is called to obtain multimodal similarity; Step S237: Call the XGBClassifier() function to train multimodal similarity and generate a mapping relationship graph based on the set cross-graph relationship constraints.

7. A method for verifying multi-source data for green certificate issuance according to claim 6, characterized in that: The step S3 specifically includes the following steps: Step S31: Obtaining the geographic grid code according to the mapping relationship map, and combining it with the obtained power generation entity ID to generate a composite primary key; Step S32: constructing an inter-provincial characteristic matrix based on the set inter-provincial output correlation coefficient, geographical attenuation factor, and constructed composite primary key; Step S33: Based on the inter-provincial feature matrix and the set contamination parameters, the Isolation Forest() function is called to determine whether the multi-source data of inter-provincial green certificate issuance is abnormal. If abnormal, an alarm is generated, isolation is carried out, and blockchain evidence storage is triggered; otherwise, step S34 is executed; Step S34: Call the GTA network to calculate the edge attention weight of the abnormal transaction path. If the edge attention weight is less than the set weight threshold, the multi-source data of the inter-provincial green certificate issuance is judged to be suspicious and transmitted to the manual review queue, and the related transactions are frozen at the same time; otherwise, execute step S35; Step S35: Call the LSTM-Attention network to predict the power generation in the next period. If the power generation is less than the deviation threshold, log it and generate a data quality report.

8. A verification system for multi-source data for green certificate issuance, used to execute the verification method for multi-source data for green certificate issuance as claimed in claim 1; characterized in that: It includes multi-source data acquisition unit, mapping unit and verification unit; The multi-source data acquisition unit is used to collect multi-source data for green certificate issuance based on the constructed multi-source heterogeneous data interface layer; The mapping unit is used to establish a data warehouse based on distributed storage and call the project information mapping algorithm to associate the multi-source data of green certificate issuance across systems; The verification unit is used to verify the associated inter-provincial green certificate issuance multi-source data based on the set intelligent verification engine fusion rules and the constructed machine learning model.

9. An electronic device, characterized in that: It includes a memory and a processor; a computer program is stored on the memory; when the computer program is executed on the processor, the verification method for multi-source data for green certificate issuance as described in any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer instructions; when the computer instructions are executed on the electronic device as described in claim 9, the verification method for multi-source data for green certificate issuance as described in any one of claims 1-7 is implemented.

Images