Multi-dimensional dynamic evaluation system and method for electric power prediction service provider

Abstract

The invention provides a multi-dimensional dynamic evaluation system and method for electric power prediction service providers, and the system comprises a data fusion perception module which is used for obtaining multi-modal data from a plurality of electric power prediction service providers; the self-adaptive dynamic evaluation module is used for comprehensively evaluating the service provider based on the multi-modal data; and the block chain evidence storage and excitation module is used for carrying out credible evidence storage on the evaluation result and the evaluation process, and executing token excitation based on the stored evaluation result. The system further comprises a data capitalization module which is used for binding a prediction data packet of a service provider with a stored evaluation result to generate prediction data NFT; the visualization and interaction module is used for providing an operation interface and an interaction interface and visually displaying the evaluation result and related data; and the preferential correction and credible issuing module is used for receiving a manual preferential correction instruction, generating and issuing a final prediction file, and uploading the instruction and the hash value of the prediction file to the block chain for evidence storage.

Description

Technical Field

[0001] This invention relates to the field of power system and energy management technology, and in particular to a multi-dimensional dynamic evaluation system and method for power forecasting service providers. Background Technology

[0002] In new energy power generation plants such as wind farms and photovoltaic power stations, power forecasting is crucial for the stable operation of the power grid and the economic benefits of the power plant. Currently, these plants typically use power forecasting data provided by third-party forecasting service providers. To select the service provider with the highest forecasting accuracy, the traditional approach is to deploy multiple power forecasting systems at each plant, with each system accessing forecasting data from one service provider. Then, manual or simple comparative analysis is performed within each plant. This traditional method has the following significant drawbacks: 1. High construction costs: Each site requires repeated investment to build multiple sets of hardware and software systems. When a company has dozens or even hundreds of sites, the total construction cost is extremely high.

[0003] 2. Inconsistent evaluation standards: Each site conducts independent evaluations, which may use different evaluation indicators or calculation methods. This leads to inconsistent evaluation results for the same service provider across different sites, making it impossible to form a unified and objective assessment at the company level.

[0004] 3. Low management efficiency: Operation and maintenance personnel need to log in to the systems of each site separately to view the forecast results. The data is scattered, and the comparative analysis work is cumbersome, making it difficult to quickly and macroscopically grasp the overall performance of all service providers in all sites.

[0005] 4. Single evaluation dimension, ignoring data value: Existing evaluation systems rely heavily on static accuracy indicators (such as RMSE, MAE), ignoring the shape of the prediction curve, uncertainty, and performance in key scenarios, and cannot fully reflect the true value of the prediction data; at the same time, the lack of an effective real-time incentive mechanism makes it difficult to motivate service providers to continuously optimize.

[0006] 5. Trust dependence of centralized systems: Even if a centralized evaluation system is built, the fairness of its data and results depends entirely on the credibility of the centralized system operator. There is a risk that the data may be tampered with or forged, the evaluation process is not transparent, and it is easy to cause disputes.

[0007] Therefore, there is an urgent need for a predictive service provider evaluation and data circulation solution that can significantly reduce costs, achieve intelligent and comprehensive evaluation, and ensure a transparent and reliable process. Summary of the Invention

[0008] The purpose of this invention is to provide a multi-dimensional dynamic evaluation system and method for power forecasting service providers, aiming to solve the above-mentioned problems in the prior art.

[0009] This invention provides a multi-dimensional dynamic evaluation system for power forecasting service providers, comprising: The data fusion and perception module is connected to the adaptive dynamic evaluation module and is used to acquire multimodal data from multiple power forecasting service providers and transmit the multimodal data to the adaptive dynamic evaluation module. An adaptive dynamic evaluation module, connected to the data fusion perception module and the blockchain evidence storage and incentive module, is used to comprehensively evaluate service providers based on the multimodal data and send the evaluation results to the blockchain evidence storage and incentive module. The blockchain notarization and incentive module is equipped with smart contracts and is connected to the adaptive dynamic evaluation module. It is used to reliably notarize the evaluation results and evaluation process, and to execute token incentives based on the notarized evaluation results.

[0010] This invention provides a multi-dimensional dynamic evaluation method for power forecasting service providers, including: The data fusion and perception module acquires multimodal data from multiple power forecasting service providers and transmits the multimodal data to the adaptive dynamic evaluation module. The adaptive dynamic evaluation module performs a comprehensive evaluation of the service provider based on the multimodal data, and sends the evaluation results to the blockchain notarization and incentive module. The evaluation results and process are reliably stored through a blockchain-based notarization and incentive module, and token incentives are executed based on the stored evaluation results.

[0011] This invention also provides an electronic device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program, when executed by the processor, implements the steps of the above-described multidimensional dynamic evaluation method for power forecasting service providers.

[0012] This invention also provides a computer-readable storage medium storing an information transmission implementation program, which, when executed by a processor, implements the steps of the aforementioned multi-dimensional dynamic evaluation method for power forecasting service providers.

[0013] The following beneficial effects can be achieved by adopting the embodiments of the present invention: The embodiments of the present invention provide a trusted ecosystem for power forecasting service providers based on blockchain and multimodal adaptive evaluation. The system builds a centralized platform that integrates advanced artificial intelligence evaluation models and blockchain trust mechanisms, realizes a comprehensive evaluation of forecasting service providers in a low-cost, highly reliable, multi-dimensional and intelligent manner, and promotes the efficient circulation and value realization of high-quality forecasting data. Attached Figure Description

[0014] To more clearly illustrate the technical solutions in one or more embodiments of this specification or in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this specification. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0015] Figure 1 This is a schematic diagram of a multi-dimensional dynamic evaluation system for power forecasting service providers according to an embodiment of the present invention; Figure 2 This is a flowchart of the multi-dimensional dynamic evaluation method for power forecasting service providers according to an embodiment of the present invention. Detailed Implementation

[0016] To enable those skilled in the art to better understand the technical solutions in one or more embodiments of this specification, the technical solutions in one or more embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this specification, and not all of the embodiments. Based on one or more embodiments of this specification, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of this document.

[0017] System Implementation Examples According to embodiments of the present invention, a multi-dimensional dynamic evaluation system for power forecasting service providers is provided. Figure 1 This is a schematic diagram of a multi-dimensional dynamic evaluation system for power forecasting service providers according to an embodiment of the present invention, such as... Figure 1 As shown, the multi-dimensional dynamic evaluation system for power forecasting service providers according to an embodiment of the present invention specifically includes: The data fusion perception module 10 is connected to the adaptive dynamic evaluation module and is used to acquire multimodal data from multiple power forecasting service providers and transmit the multimodal data to the adaptive dynamic evaluation module. The adaptive dynamic evaluation module 12, connected to the data fusion perception module and the blockchain evidence storage and incentive module, is used to comprehensively evaluate service providers based on the multimodal data and send the evaluation results to the blockchain evidence storage and incentive module. Specifically, it is used for: Based on the multimodal data, the original scores of each service provider on multiple evaluation indicators are calculated using a multi-dimensional evaluation indicator library; wherein, the multimodal data includes power prediction data, actual power data of the power station, and multi-source heterogeneous context data; the multiple evaluation indicators include at least: a shape similarity indicator for evaluating the similarity of prediction curves, an uncertainty quantification indicator for evaluating the sharpness and calibration of probabilistic predictions, and a key scenario performance indicator for evaluating prediction performance in specific scenarios. The weights of the multiple evaluation indicators are dynamically assigned using the game theory Shapley value method or a deep learning model. The original scores are weighted based on the weights to obtain the comprehensive score and ranking of each service provider.

[0018] The blockchain evidence storage and incentive module 14 is equipped with a smart contract and is connected to the adaptive dynamic evaluation module. It is used to reliably store the evaluation results and evaluation process, and to execute token incentives based on the stored evaluation results. Specifically, the smart contract includes: The dynamic evaluation contract encapsulates evaluation logic and is used to automatically execute a comprehensive evaluation process for power forecasting service providers, and to store the evaluation results and process data on the blockchain for evidence. The transaction contract, connected to the dynamic evaluation contract, is used for on-chain clearing and token transfer of predictive data transactions between the venue operator and the service provider. An incentive contract, connected to the dynamic evaluation contract, is used to automatically issue incentive tokens to top-ranked service providers based on the evaluation results stored on the blockchain.

[0019] The system further includes: The data assetization module, connected to the blockchain notarization and incentive module, is used to bind the service provider's prediction data package with the notarized evaluation results to generate prediction data NFTs. The visualization and interaction module is connected to the blockchain notarization and incentive module and the optimization correction and trusted distribution module. It is used to provide users with a user-friendly operation interface and interaction interface, and to visualize the evaluation results and related data. The evaluation results and related data specifically include: a comprehensive score and ranking list of each power forecasting service provider, a multi-dimensional capability radar chart for intuitively comparing the capabilities of each service provider in different evaluation dimensions, a Shapley value analysis dashboard for revealing the weight of key indicators in this evaluation, and blockchain-based evidence data. The optimization and reliable distribution module is connected to the blockchain evidence storage and incentive module and the visualization and interaction module. It is used to receive manual optimization and modification instructions from the visualization and interaction module, generate and distribute the final prediction file, and upload the instructions and the hash value of the final prediction file to the blockchain evidence storage and incentive module for evidence storage.

[0020] The following describes the above-mentioned technical solution of the present invention in detail with reference to the specific situation of the multi-dimensional dynamic evaluation system for power forecasting service providers in the embodiments of the present invention.

[0021] This invention proposes a trusted ecosystem for power forecasting service providers based on blockchain and multimodal adaptive evaluation. The system includes: a multimodal data fusion and sensing module for accessing forecast data, actual power data, and multi-source heterogeneous data such as meteorological and dispatching demand data; an adaptive dynamic evaluation engine employing a multi-dimensional indicator library including shape similarity, uncertainty quantification, and key scenario performance, and dynamically allocating indicator weights based on game theory Shapley values ​​or deep learning models to achieve intelligent and fair evaluation; a blockchain core layer that automates and reliably stores the evaluation process through smart contracts, and supports token-based data trading and incentives; a data assetization module that binds forecast data and trusted evaluation data to NFTs (non-fungible tokens) to promote data value circulation; and a visualization interface and a selective distribution module. This invention solves the problems of high construction costs, opaque evaluation, single dimensions, and lack of effective incentives in existing technologies, achieving low-cost, highly reliable, multi-dimensional intelligent evaluation of forecasting service providers and efficient circulation of high-quality data. Specifically, the system architecture mainly includes the following core modules: 1. Multimodal Data Fusion Sensing Module: This module is the data foundation of the system. It centrally integrates power forecast data from multiple forecasting service providers for various power plants under the company, as well as the actual power data of the power plants. It also integrates multi-source contextual data, including but not limited to: refined numerical weather prediction (NWP) data (such as irradiance, wind speed, wind direction, cloud cover, and temperature), grid dispatch demand-side signals (such as peak-shaving instructions, frequency regulation requirements, and real-time electricity prices), and power plant equipment status data (such as the health status of inverters and wind turbines, which may affect power generation potential). This provides a reliable data foundation for the system's subsequent in-depth, scenario-based intelligent evaluation.

[0022] 2. Adaptive Dynamic Evaluation Engine: This module is the intelligent brain of the system and is responsible for conducting in-depth evaluations.

[0023] (1) Multi-dimensional evaluation index library: Based on traditional accuracy indicators (such as RMSE, MAE), the following are introduced: A. Shape similarity index: Using algorithms such as Dynamic Time Warping (DTW), the similarity between the predicted curve and the actual curve in terms of shape and phase is evaluated. It is especially suitable for evaluating the ability to capture key events such as hill climbing.

[0024] B. Quantitative Indicators for Predictive Uncertainty: This module requires service providers to provide probability predictions or uncertainty intervals (such as quantile predictions). It evaluates the sharpness (the degree of concentration of the interval, the narrower the interval the better) and calibration (whether the probability of the actual value falling within the interval is accurate) of the prediction, thereby measuring the reliability of the prediction.

[0025] C. Key Scenario Performance Metrics: The system can automatically identify scenarios such as extreme weather, peak load periods, and rapid ramp-up events, and evaluate the service provider's predictive performance at these times separately.

[0026] (2) Weighting model: The weights of each indicator are dynamically determined through two optimization methods: A. Game Theory-Driven Approach: The evaluation process is modeled as a cooperative game, using Shapley Value to calculate the marginal contribution of each evaluation indicator in differentiating service provider capabilities. Indicators with greater contributions receive higher weights within the current evaluation period, thus automatically focusing on the most discriminative evaluation dimensions. The specific implementation process is as follows: S1: Modeling Participants: Not the forecasting service provider, but each evaluation metric (such as accuracy RMSE, shape DTW, uncertainty sharpness, etc.), assuming there are M metrics.

[0027] Coalition: Any combination of indicators is a coalition, for example, {precision} is one coalition, and {precision, shape} is another coalition.

[0028] Alliance revenue: Define a revenue function This is used to quantify the system's ability to distinguish between good and bad service providers when only the set of metrics in Alliance S is used.

[0029] S2: Profit Function Definition The payoff function needs to be able to measure "discrimination," and an effective and computable definition is: =The standard deviation of all service provider scores calculated using only the indicators in Alliance S; the larger the standard deviation, the more significant the differences in scores between service providers, meaning the stronger the discriminative power of this indicator combination. If all service providers have the same score, the standard deviation is 0, meaning this indicator combination cannot distinguish between good and bad.

[0030] S3: Calculate the Shapley value for each indicator. Input: The original score matrix of all N service providers on M indicators within a period (e.g., one day).

[0031] The formula for calculating the Shapley value is: ; For each indicator i, iterate through all alliances S (where S is a subset of M indicators) that do not contain i. Calculate the marginal benefit (i.e., the marginal return) that comes with indicator i joining alliance S. Then, the marginal revenue of all possible alliances S is weighted and averaged; this average is the Shapley value of indicator i. .in, It is a weight in combinatorial mathematics, used to ensure that all possible permutations are equally likely.

[0032] S4: Weighting: Normalize all Shapley values ​​so that their sum is 1. ; in, It refers to the dynamic weight assigned to the i-th indicator in the current period.

[0033] Output: Using dynamic weights Calculate the final overall score for each service provider and rank them.

[0034] B. Deep Learning-Driven Approach: A deep learning network (such as a model with an attention mechanism) is constructed, using multimodal contextual data as input. The network is trained to maximize the correlation between the evaluation results and the final economic benefits of the power station (such as power generation revenue and ancillary service compensation). This network can proactively and dynamically adjust the weights of various indicators based on future grid conditions and market realities. The specific implementation process is as follows: S1: Network Structure Design Input layer: Receives multimodal context data, such as: forecasted weather data (sequence) for the next 24 hours; electricity market price signals (sequence) for the next 24 hours; peak-shaving / frequency regulation demand plans released by the power grid; date type (weekday / holiday).

[0035] Core layer: A neural network using an attention mechanism (such as the encoder part of a Transformer or an LSTM / GRU with attention). The attention mechanism allows the model to automatically and dynamically assign different importance (i.e., weights) to different input features (corresponding to different evaluation metrics).

[0036] Output layer: A softmax layer with an output dimension of M (the number of evaluation metrics). The output value is the dynamic weight of each metric. And all The sum is 1.

[0037] S2: Training process (conducted offline) a. Prepare training data: Features: Historical multimodal contextual data.

[0038] Tags: This is an unsupervised or self-supervised process that does not require manual annotation. The training objective of this invention is to ensure that the service provider rankings calculated based on the weights of the network output best predict the future actual economic benefits of the site.

[0039] b. Define the loss function: For a piece of historical data, the features are input into the network to obtain the index weights w; The weight w is used to weight the scores of each indicator of all service providers at that time, so as to obtain the comprehensive score of each service provider and rank them. The predicted data from the top-ranked service provider was considered the data selected at that time; Calculate the simulated economic benefits that using the forecast data from this "champion" service provider could bring to the site. For example: ; The above formula rewards accurate predictions of power generation (participation in market transactions) and penalizes prediction errors (leading to balancing costs).

[0040] The loss function L = -revenue. The objective of this embodiment of the invention is to minimize this loss, that is, to maximize the simulated economic benefit.

[0041] c. Model training: Optimization algorithms such as gradient descent are used to repeatedly adjust network parameters; Through this process, the network gradually learns that when a certain combination of weather and market prices occurs (such as high electricity prices and sunny weather tomorrow afternoon), higher weight should be given to peak prediction accuracy and shape similarity to ensure that the service provider that can most accurately predict the peak generation is selected, thereby maximizing electricity sales revenue.

[0042] Preferably, in practical systems, the two methods described above can coexist or even be integrated. For example, the results of the Shapley value method can be used as an auxiliary input to a deep learning model, or operators can choose which method of weight generation to use based on different management objectives, thereby enhancing the scientific nature and flexibility of the system.

[0043] 3. Blockchain Core Layer: This module is the cornerstone of the system's trust, providing transparent, tamper-proof, and reliable protection for the entire system.

[0044] (1) Distributed ledger: Stores hash values ​​of key data (such as prediction data hash, evaluation result hash, weight vector hash), as well as token circulation records. Original large files can be stored in decentralized storage networks such as IPFS, and on-chain anchoring and verification can be performed through hash values.

[0045] (2) Smart contract module, including: A. Dynamic Evaluation Contract: Encapsulates the logic of the aforementioned adaptive dynamic evaluation engine or its output interface. The evaluation process is executed automatically by the code, the rules are public and transparent, and the results are recorded on the blockchain after consensus is reached, making them tamper-proof.

[0046] B. Trading Contract: Market transactions that manage forecast data. The venue pays digital tokens to purchase forecast data from a specific service provider. Once the data is confirmed and issued, the contract is automatically settled, realizing the instant transfer of value.

[0047] C. Incentive Contract: Based on the trusted evaluation results on the blockchain, incentive tokens are automatically issued to top-ranked or high-performing service providers, creating a positive economic incentive cycle that requires no human intervention.

[0048] 4. Data Assetization and NFT Issuance Module: This module acts as a value amplifier for the system. It binds the prediction data packages provided by service providers with their blockchain-certified, multi-dimensional evaluation reports, jointly creating unique prediction data NFTs. This makes each piece of prediction data a digital asset with proof of quality and scarcity, which can be traded on the secondary market, greatly unlocking and demonstrating the potential value of prediction data.

[0049] 5. Visualization and Application Interaction Interface: Provides users with an intuitive and transparent interactive window, offering not only standard curve comparisons and ranking displays, but also: (1) Multidimensional capability radar chart: intuitively displays the service provider's capability boundaries in each evaluation dimension; (2) Shapley value analysis dashboard: reveals which indicators played a key role in this evaluation; (3) Blockchain explorer: Allows any participant to verify all on-chain evidence records in real time; (4) Predictive Data NFT Marketplace: A platform for browsing and trading high-quality predictive data assets.

[0050] 6. Optimal Correction and Trusted Distribution Module: The final output stage of the system. Based on the trusted, multi-dimensional evaluation results on the blockchain, the site administrator selects the best prediction source. All manual correction operations are recorded and generated with operation hashes, which are then stored on the blockchain to ensure the traceability and non-repudiation of the operations. The hash value of the final distributed prediction file is also stored on the blockchain, forming a complete and trusted closed loop from evaluation, selection, correction to distribution.

[0051] The system proposed in this invention can be deployed on a cloud platform or a private data center. The off-chain module adopts a microservice architecture, is developed using languages ​​such as Python / Java, and handles data access, preprocessing, and complex AI model calculations; the blockchain core layer can be built on an enterprise-grade blockchain platform such as FISCO BCOS or Hyperledger Fabric; smart contracts are written in languages ​​such as Solidity / Go; and the front-end interface is implemented using frameworks such as Vue / React.

[0052] When applying the system, the data entry point is first standardized through the data access module; then, smart contracts are deployed and configured; historical data is used to train the deep learning weight allocation model; after the system is put into operation, the data collection, on-chain and off-chain collaborative computing, evaluation, incentive and data release processes are automatically executed daily; administrators log in to the system through a browser to complete the supervision, selection and distribution work.

[0053] Method Implementation Examples According to embodiments of the present invention, a multi-dimensional dynamic evaluation method for power forecasting service providers is provided. Figure 2 This is a flowchart of the multi-dimensional dynamic evaluation method for power forecasting service providers according to an embodiment of the present invention, such as... Figure 2 As shown, the multi-dimensional dynamic evaluation method for power forecasting service providers according to an embodiment of the present invention specifically includes: Step S201: Obtain multimodal data from multiple power forecasting service providers through the data fusion sensing module, and transmit the multimodal data to the adaptive dynamic evaluation module.

[0054] Step S202 involves using an adaptive dynamic evaluation module to comprehensively evaluate the service provider based on the multimodal data, and then sending the evaluation results to the blockchain notarization and incentive module. Specifically, this includes: The adaptive dynamic evaluation module calculates the original scores of each service provider on multiple evaluation indicators based on the multimodal data and a multi-dimensional evaluation indicator library. The multimodal data includes power prediction data, actual power data of power plants, and multi-source heterogeneous context data. The multiple evaluation indicators include at least: a shape similarity indicator for evaluating the similarity of prediction curves, an uncertainty quantification indicator for evaluating the sharpness and calibration of probabilistic predictions, and a key scenario performance indicator for evaluating prediction performance in specific scenarios. The weights of the multiple evaluation indicators are dynamically assigned using the game theory Shapley value method or a deep learning model. The original scores are weighted based on the weights to obtain the comprehensive score and ranking of each service provider.

[0055] Step S203: The evaluation results and evaluation process are reliably stored through the blockchain notarization and incentive module, and token incentives are executed based on the notarized evaluation results.

[0056] The method further includes: The data assetization module binds the service provider's prediction data package with the certified evaluation results to generate prediction data NFTs. The visualization and interaction modules provide users with a user-friendly interface and interactive interface, and visualize the evaluation results and related data. The optimization and trusted delivery module receives manual optimization and modification instructions from the visualization and interaction module, generates and delivers the final prediction file, and uploads the instructions and the hash value of the final prediction file to the blockchain notarization and incentive module for notarization.

[0057] The embodiments of the present invention are method embodiments corresponding to the system embodiments described above. The specific operations of each step can be understood by referring to the description of the system embodiments, and will not be repeated here.

[0058] In summary, compared with the prior art, the beneficial effects of the embodiments of the present invention are as follows: 1. Revolutionary depth and intelligence in evaluation: Through a multi-dimensional indicator library and an adaptive weight model, the system can identify service providers that excel in accuracy, shape, reliability, and key scenario performance. The evaluation results are more scientific and fair, and can effectively drive technological progress in the industry.

[0059] 2. An irrefutable trust mechanism has been established: Blockchain technology ensures transparency and immutability throughout the entire process from data source and evaluation to transaction and incentive, creating a trust environment that does not require endorsement from a central institution, thus fundamentally eliminating disputes.

[0060] 3. Created a vibrant data value ecosystem: Through "token economy" and "data NFTization", predictive data is transformed into digital assets that can be flexibly traded and invested in, providing service providers with immediate and diversified economic returns and stimulating market vitality.

[0061] 4. Achieved the unity of intensification and intelligence: While solving all site needs with a single system and significantly reducing construction costs, it provides profound value far exceeding the capabilities of distributed systems through advanced technology, achieving ultimate optimization of cost and efficiency.

[0062] Device Example 1 This invention provides an electronic device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, it performs the steps described in the method embodiment.

[0063] Device Example 2 This invention provides a computer-readable storage medium storing an information transmission implementation program, which, when executed by a processor, performs the steps described in the method embodiment.

[0064] The computer-readable storage media described in this embodiment include, but are not limited to, ROM, RAM, disk, or optical disk.

[0065] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A multi-dimensional dynamic evaluation system for power forecasting service providers, characterized in that, include: The data fusion and perception module is connected to the adaptive dynamic evaluation module and is used to acquire multimodal data from multiple power forecasting service providers and transmit the multimodal data to the adaptive dynamic evaluation module. An adaptive dynamic evaluation module, connected to the data fusion perception module and the blockchain evidence storage and incentive module, is used to comprehensively evaluate service providers based on the multimodal data and send the evaluation results to the blockchain evidence storage and incentive module. The blockchain notarization and incentive module is equipped with smart contracts and is connected to the adaptive dynamic evaluation module. It is used to reliably notarize the evaluation results and evaluation process, and to execute token incentives based on the notarized evaluation results.

2. The system according to claim 1, characterized in that, The system further includes: The data assetization module, connected to the blockchain notarization and incentive module, is used to bind the service provider's prediction data package with the notarized evaluation results to generate prediction data NFTs. The visualization and interaction module is connected to the blockchain notarization and incentive module and the optimization correction and trusted distribution module. It is used to provide users with a user-friendly operation interface and interaction interface, and to visualize the evaluation results and related data. The optimization and reliable distribution module is connected to the blockchain evidence storage and incentive module and the visualization and interaction module. It is used to receive manual optimization and modification instructions from the visualization and interaction module, generate and distribute the final prediction file, and upload the instructions and the hash value of the final prediction file to the blockchain evidence storage and incentive module for evidence storage.

3. The system according to claim 1, characterized in that, The adaptive dynamic evaluation module is specifically used for: Based on the multimodal data, the original scores of each service provider on multiple evaluation indicators are calculated using a multi-dimensional evaluation indicator library; wherein, the multimodal data includes power prediction data, actual power data of the power station, and multi-source heterogeneous context data; the multiple evaluation indicators include at least: a shape similarity indicator for evaluating the similarity of prediction curves, an uncertainty quantification indicator for evaluating the sharpness and calibration of probabilistic predictions, and a key scenario performance indicator for evaluating prediction performance in specific scenarios. The weights of the multiple evaluation indicators are dynamically assigned using the game theory Shapley value method or a deep learning model. The original scores are weighted based on the weights to obtain the comprehensive score and ranking of each service provider.

4. The system according to claim 1, characterized in that, The smart contract specifically includes: The dynamic evaluation contract encapsulates evaluation logic and is used to automatically execute a comprehensive evaluation process for power forecasting service providers, and to store the evaluation results and process data on the blockchain for evidence. The transaction contract, connected to the dynamic evaluation contract, is used for on-chain clearing and token transfer of predictive data transactions between the venue operator and the service provider. An incentive contract, connected to the dynamic evaluation contract, is used to automatically issue incentive tokens to top-ranked service providers based on the evaluation results stored on the blockchain.

5. The system according to claim 2, characterized in that, The evaluation results and related data specifically include: a comprehensive score and ranking list of each power forecasting service provider, a multi-dimensional capability radar chart for intuitively comparing the strengths and weaknesses of each service provider in different evaluation dimensions, a Shapley value analysis dashboard for revealing the weights of key indicators in this evaluation, and blockchain-based evidence data.

6. A multi-dimensional dynamic evaluation method for power forecasting service providers, characterized in that, include: The data fusion and perception module acquires multimodal data from multiple power forecasting service providers and transmits the multimodal data to the adaptive dynamic evaluation module. The adaptive dynamic evaluation module performs a comprehensive evaluation of the service provider based on the multimodal data, and sends the evaluation results to the blockchain notarization and incentive module. The evaluation results and process are reliably stored through a blockchain-based notarization and incentive module, and token incentives are executed based on the stored evaluation results.

7. The method according to claim 6, characterized in that, The method further includes: The data assetization module binds the service provider's prediction data package with the certified evaluation results to generate prediction data NFTs. The visualization and interaction modules provide users with a user-friendly interface and interactive interface, and visualize the evaluation results and related data. The optimization and trusted delivery module receives manual optimization and modification instructions from the visualization and interaction module, generates and delivers the final prediction file, and uploads the instructions and the hash value of the final prediction file to the blockchain notarization and incentive module for notarization.

8. The method according to claim 6, characterized in that, The comprehensive evaluation of service providers based on the multimodal data through the adaptive dynamic evaluation module specifically includes: The adaptive dynamic evaluation module calculates the original scores of each service provider on multiple evaluation indicators based on the multimodal data and a multi-dimensional evaluation indicator library. The multimodal data includes power prediction data, actual power data of power plants, and multi-source heterogeneous context data. The multiple evaluation indicators include at least: a shape similarity indicator for evaluating the similarity of prediction curves, an uncertainty quantification indicator for evaluating the sharpness and calibration of probabilistic predictions, and a key scenario performance indicator for evaluating prediction performance in specific scenarios. The weights of the multiple evaluation indicators are dynamically assigned using the game theory Shapley value method or a deep learning model. The original scores are weighted based on the weights to obtain the comprehensive score and ranking of each service provider.

9. An electronic device, characterized in that, include: The memory, the processor, and the computer program stored in the memory and executable on the processor, wherein the computer program, when executed by the processor, implements the steps of the multidimensional dynamic evaluation method for power forecasting service providers as described in any one of claims 6-8.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores an information transmission implementation program, which, when executed by a processor, implements the steps of the multi-dimensional dynamic evaluation method for power forecasting service providers as described in any one of claims 6-8.

Images