A Method for Constructing a Trust Mechanism for Electricity Market Transactions Based on Blockchain

Abstract

This invention discloses a method for constructing a trust mechanism for electricity market transactions based on blockchain, relating to the field of electricity market transaction technology. The specific steps of this construction method are as follows: S100, data collection and feature fusion: collecting electricity user equipment operation data, electricity consumption behavior pattern data, power grid interaction data, as well as electricity market transaction price fluctuation data, historical default records, and power grid operation status data. By utilizing the distributed ledger characteristics of blockchain, this invention achieves encrypted storage and tamper-proofing of multi-dimensional information such as electricity user equipment operation data, electricity consumption behavior pattern data, power grid interaction data, and electricity market transaction price fluctuation data, effectively preventing data forgery and tampering. At the same time, by using fuzzy membership functions to create an initial credit scoring model and combining it with a real-time monitoring and scoring update mechanism, this invention can dynamically adjust the credit scores of trading entities, ensuring the accuracy and timeliness of credit assessment.

Description

Technical Field

[0001] This invention relates to the field of electricity market trading technology, specifically a method for constructing a trust mechanism for electricity market trading based on blockchain. Background Technology

[0002] With the transformation of the global energy structure and the rapid development of smart grids, electricity market transactions are becoming increasingly complex and diversified. Traditional electricity market transaction models can no longer meet the demands of modern power systems for an efficient, transparent, and trustworthy transaction environment. Against this backdrop, blockchain technology, with its decentralized, immutable, transparent, and traceable characteristics, provides a new solution for building a trust mechanism for electricity market transactions. Blockchain technology records transaction information through a distributed ledger, ensuring the authenticity and integrity of data, and providing a more secure and reliable trust foundation for electricity market transactions. Therefore, applying blockchain technology to the construction of a trust mechanism for electricity market transactions has become a research hotspot and technological development trend in the current energy field.

[0003] In traditional electricity market transactions, the trust mechanism mainly relies on centralized third-party institutions for credit assessment and supervision. However, this model has many shortcomings. First, centralized institutions may face the risk of single point of failure. Once a problem occurs, the trust foundation of the entire trading system will be severely affected. Second, traditional credit building methods are often based on limited historical data and static assessment models, which are difficult to accurately reflect the real-time credit status of trading entities, resulting in lagging and inaccurate credit assessment results. In addition, the traditional trading model suffers from poor data flow and serious information asymmetry, which increases transaction costs and risks. Therefore, traditional technologies are clearly insufficient in ensuring the transparency, credibility, and risk management of electricity market transactions, and new technological means are urgently needed to improve them.

[0004] Therefore, developing a blockchain-based trust mechanism for electricity market transactions not only enhances the transparency and credibility of electricity market transactions but also effectively reduces transaction risks and promotes the healthy and stable development of the electricity market. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of existing technologies and provide a method for constructing a trust mechanism for electricity market transactions based on blockchain. By collecting multi-dimensional data and performing feature fusion, a credit scoring model is created using fuzzy membership functions to achieve initial credit scoring and real-time dynamic updates of transaction entities. At the same time, a risk assessment model is constructed to quantify transaction risks and preset risk warning thresholds to achieve intelligent early warning. This effectively improves the transparency, credibility, and risk management capabilities of electricity market transactions, and provides a strong guarantee for the stable operation of the electricity market.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: On the one hand, a method for constructing a blockchain-based trust mechanism for electricity market transactions, the specific steps of which are as follows:

[0007] S100, Data Acquisition and Feature Fusion: Collects power user equipment operation data, electricity consumption behavior pattern data, power grid interaction data, as well as power market transaction price fluctuation data, historical default records, and power grid operation status data, performs preprocessing and feature extraction, and merges them into a comprehensive feature vector, which is then encrypted and stored in the blockchain distributed ledger.

[0008] S200, Credit Modeling and Initial Assessment: An initial credit scoring model is created based on a fuzzy membership function, and the initial credit scoring model is trained and optimized using a comprehensive feature vector. The initial credit score of the transaction entity is calculated through the optimized model and recorded on the blockchain.

[0009] S300, Real-time Monitoring and Scoring Update: Real-time tracking of the operating status of power generation equipment and user electricity consumption behavior. Once equipment failure or abnormal electricity consumption is detected, the latest comprehensive feature vector in the blockchain is retrieved, combined with the credit score and feature change amount at the previous moment, and the real-time credit score of the transaction subject is updated through the real-time credit scoring algorithm and synchronized to the blockchain.

[0010] S400, Risk Modeling and Quantitative Assessment: Analyzes market data stored on the blockchain, and based on the analysis results, constructs a risk assessment model using the real-time credit scores of the trading entities and market price fluctuation data as inputs, and calculates the risk value of each electricity transaction;

[0011] S500, Risk Warning and Smart Trigger: Based on market conditions, a risk warning threshold is preset. When the trading risk value reaches the threshold, the probability of triggering the warning is calculated. When the probability exceeds the preset threshold, the smart contract is automatically triggered to send warning information to relevant trading parties and regulatory agencies.

[0012] Furthermore, in S100, the acquisition methods and content of various types of data in data acquisition and feature fusion are as follows:

[0013] Power user equipment operation data: By deploying temperature sensors, vibration sensors, and current sensors on power user equipment, the operating parameters of the equipment, such as temperature, vibration frequency, and current intensity, are collected in real time; through the equipment's built-in monitoring module, data on equipment operating time, number of start-stop cycles, and fault code status are obtained;

[0014] Electricity consumption behavior pattern data: Real-time active power, reactive power, and power factor data of users are collected using smart meters;

[0015] Grid interaction data: Obtain voltage stability data during power transmission from the grid-side monitoring system and record power interaction data between the grid and users;

[0016] Electricity market transaction price fluctuation data: Access the electricity market trading platform to obtain electricity transaction prices in real time for different trading periods; collect historical transaction price data and statistically analyze the price fluctuation range, fluctuation frequency, and price change trend.

[0017] Collect historical default records: retrieve the default history of the trading entities from the power market trading management system, and record the data of the default time, defaulting party name, default type, and default handling result;

[0018] Power grid operation status data: acquire operating indicators such as overall power grid load rate, line transmission power limit, and transformer load rate; monitor power quality data such as power grid frequency deviation, harmonic content, and three-phase imbalance, as well as the on / off status information of power grid equipment.

[0019] Furthermore, in S100, during data acquisition and feature fusion, a multi-dimensional feature fusion algorithm is used to fuse multi-source data into a comprehensive feature vector. Let the feature vector of the device operation data be... The feature vector of electricity consumption behavior pattern data is The feature vector of power grid interaction data is The feature vector of market-related data is Integrate them into a comprehensive feature vector The calculation formula is: ,in, Weighting coefficients are pre-defined based on the importance of the data, and , This indicates a feature fusion operation, specifically the addition of corresponding elements.

[0020] Furthermore, in S200, the initial credit scoring model is created by combining fuzzy membership functions in credit modeling and initial assessment. Let the fuzzy membership function be... This is used to measure the impact of data features on credit scores, with the initial credit score being... The calculation formula is: ,in, For the comprehensive feature vector The first in One element, for The fuzzy membership degree belonging to high-credit characteristics is determined by a fuzzy rule table. It is the feature vector of equipment operation data. The number of elements, It is a feature vector of electricity consumption behavior pattern data. The number of elements, It is the feature vector of power grid interaction data. The number of elements, It is a feature vector of market-related data. The number of elements.

[0021] Furthermore, in S200, the training steps for the initial credit scoring model in credit modeling and initial assessment are as follows:

[0022] (1) Collect historical comprehensive feature vectors and divide the training set and test set in a 7:3 ratio;

[0023] (2) Preset the initial parameters of the fuzzy membership function based on domain knowledge, and define the fuzzy rule table;

[0024] (3) Input the training set into the initial credit scoring model, iteratively adjust the parameters based on the training set, and use the test set to calculate the F1 score to verify the model's generalization ability. The calculation formula is as follows: Where: TP is a true positive instance, representing a correctly predicted positive sample; FP is a false positive instance, representing a incorrectly predicted positive sample; and FN is a false negative instance, representing a incorrectly predicted negative sample.

[0025] (4) When the F1 score is ≥0.85, the model meets the standard and the generalization ability of the model is considered to be up to standard, and training is stopped; if it does not meet the standard, return to step (3), readjust the fuzzy rule parameters and repeat the training and verification process until the index requirements are met.

[0026] (5) After verification, the final fuzzy rule parameters and model structure are solidified and deployed to the credit assessment module of the blockchain system for the calculation of the initial credit score of the transaction subject.

[0027] Furthermore, in S300, the real-time credit score of the transaction entity is updated through a real-time credit scoring algorithm during real-time monitoring and scoring updates. Let the real-time credit score be... The calculation formula is: ,in, This is the score from the previous moment. For the period from the last rating to the current moment The change in each characteristic This is a time decay factor that decreases as the scoring time interval increases. The coefficients are dynamically adjusted based on the degree of data anomalies. It is the feature vector of equipment operation data. The number of elements, It is a feature vector of electricity consumption behavior pattern data. The number of elements, It is the feature vector of power grid interaction data. The number of elements, It is a feature vector of market-related data. The number of elements.

[0028] Furthermore, in S400, the risk assessment model is constructed during risk model construction and evaluation, assuming the risk value is... The calculation formula is: ,in, The average credit score in the market. These are the weighting coefficients, and , Market data characteristics Risk sensitivity coefficient Current credit score It is a feature vector of market-related data. The number of elements.

[0029] Furthermore, in S500, the calculation of the early warning trigger probability in risk warning and intelligent triggering is based on the risk warning threshold being... ,when When the warning mechanism is triggered, the warning trigger probability is calculated. Let the warning trigger probability be W, and the calculation formula is: ,in, This is an adjustable parameter used to control the sensitivity to probability changes, with a value range of 0.5 to 2.0. An alert is triggered when the probability threshold of 0.7 is exceeded.

[0030] On the other hand, a blockchain-based trust mechanism construction system for electricity market transactions includes: a data acquisition and fusion module, a credit modeling and initial assessment module, a real-time monitoring and scoring update module, a risk modeling and quantitative assessment module, and a risk warning and intelligent triggering module.

[0031] The data acquisition and fusion module collects multi-dimensional data on equipment operation, electricity consumption behavior, grid interaction, market prices, default records, and grid status through sensors, smart meters, power grid monitoring systems, and market platforms. After preprocessing and feature extraction, the data is fused into a comprehensive feature vector and encrypted and stored in a blockchain distributed ledger.

[0032] The credit modeling and initial assessment module: constructs an initial credit scoring model based on fuzzy membership functions, divides the historical comprehensive feature vector into training and test sets according to a ratio, optimizes the model by iteratively adjusting parameters, calculates the initial credit score of the transaction entity using the optimized model, and records it to the blockchain;

[0033] The real-time monitoring and scoring update module: tracks the operating status of power generation equipment and user electricity consumption behavior in real time. When equipment failure or abnormal electricity consumption is detected, it retrieves the latest comprehensive feature vector of the blockchain, combines it with the credit score and feature change amount of the previous moment, updates the credit score of the transaction subject through the real-time credit scoring algorithm, and synchronizes it to the blockchain.

[0034] The risk modeling and quantitative assessment module analyzes market data stored on the blockchain, uses real-time credit scores of trading entities and market price fluctuation data as inputs, constructs a risk assessment model, and calculates the risk value of each electricity transaction.

[0035] The risk warning and intelligent triggering module: presets a risk warning threshold based on market conditions; when the transaction risk value reaches the threshold, it calculates the warning trigger probability; when the probability exceeds the preset threshold, it automatically triggers the smart contract to send warning information to relevant trading parties and regulatory agencies.

[0036] Compared with existing technologies, this blockchain-based method for building a trust mechanism for electricity market transactions has the following advantages:

[0037] I. This invention utilizes the distributed ledger characteristics of blockchain to achieve encrypted storage and immutability of multi-dimensional information, including power user equipment operation data, electricity consumption behavior pattern data, power grid interaction data, and electricity market transaction price fluctuation data. This effectively prevents data forgery and tampering. Furthermore, by using fuzzy membership functions to create an initial credit scoring model and combining it with a real-time monitoring and scoring update mechanism, this invention can dynamically adjust the credit scores of trading entities, ensuring the accuracy and timeliness of credit assessment. This blockchain-based trust mechanism construction method not only enhances the transparency and fairness of electricity market transactions but also promotes mutual trust and cooperation between trading parties, providing a strong guarantee for the healthy development of the electricity market.

[0038] Second, this invention analyzes market data stored on the blockchain and combines it with real-time credit scores and market price fluctuation data of trading entities to construct a risk assessment model. This model can accurately calculate the risk value of each electricity transaction. Based on this, a risk warning threshold is preset, and when the transaction risk value reaches the threshold, an early warning mechanism is automatically triggered to send warning information to relevant trading parties and regulatory agencies. This intelligent risk warning and triggering mechanism makes risk management in the electricity market more proactive and efficient, helps to promptly identify and respond to potential market risks, and protects the legitimate rights and interests of both trading parties. At the same time, this mechanism also promotes the intelligence and refinement of electricity market supervision, improves regulatory efficiency, and provides solid support for the stable operation of the electricity market.

[0039] Other advantages, objectives and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination or study, or may be learned from the practice of the invention. Attached Figure Description

[0040] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0041] Figure 1 A framework diagram for constructing a blockchain-based trust mechanism for electricity market transactions;

[0042] Figure 2 A flowchart for building a blockchain-based trust mechanism for electricity market transactions. Detailed Implementation

[0043] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below.

[0044] Example 1:

[0045] Please see Figure 1 , Figure 1 A framework diagram is provided for constructing a blockchain-based trust mechanism for electricity market transactions, which can be used to assess the trust in grid connection transactions between the power grid and distributed photovoltaic users.

[0046] Taking this scenario as an example, the city power grid company is responsible for managing the access of distributed energy within its jurisdiction. It signs a "Distributed Photovoltaic Power Generation Project Grid Connection Agreement" with household photovoltaic users, stipulating that users will connect surplus photovoltaic power to the grid, and the power grid company will pay the grid connection fee according to policy. Since the user's photovoltaic equipment directly affects the power quality and dispatch security of the power grid, it is necessary to dynamically evaluate the stability, compliance and transaction reliability of the equipment through a blockchain-based power market transaction trust mechanism to prevent grid risks caused by equipment failure or substandard power quality.

[0047] To address the issues in this scenario, a method for constructing a blockchain-based trust mechanism for electricity market transactions is provided, comprising the following steps:

[0048] S100, Data Acquisition and Feature Fusion: Collects power user equipment operation data, electricity consumption behavior pattern data, power grid interaction data, as well as power market transaction price fluctuation data, historical default records, and power grid operation status data. It performs preprocessing and feature extraction, and merges them into a comprehensive feature vector, which is then encrypted and stored in the blockchain distributed ledger.

[0049] Specifically, S100 includes:

[0050] Equipment operation data acquisition: Through temperature sensors and current sensors deployed on the user's photovoltaic inverter, the core operating parameters of the equipment are collected in real time: internal temperature of the inverter (reflecting the heat dissipation status), output current intensity (reflecting the power generation), cumulative running time and number of start-stop cycles (to assess equipment wear), and fault codes (such as overload and short circuit).

[0051] Electricity consumption behavior and grid connection data collection: Using smart meters and grid-side monitoring systems to obtain: active power and power factor: real-time power injected by users into the grid (negative values ​​indicate generation and grid connection) and power utilization efficiency; power quality indicators: voltage deviation, harmonic content, three-phase imbalance (reflecting whether grid-connected power meets national standards); dispatch instruction execution status: whether users adjust generation power according to grid requirements (such as limiting grid-connected power during high grid load periods).

[0052] Market and credit data collection: Access to the power market system and regulatory platform to obtain: distributed photovoltaic feed-in tariff policies (such as subsidy standards for different periods), household historical settlement records (whether the power generation was declared and subsidies were received on time), and historical default records (such as being issued rectification notices by the power grid company due to excessive harmonics).

[0053] Equipment operation data, electricity consumption behavior data, power grid interaction data, and market credit data are preprocessed into standardized features (e.g., mapping "harmonic exceedance" to a negative feature value). These features are then fused using a multi-dimensional feature fusion algorithm to generate a comprehensive feature vector that includes dimensions of equipment status, compliance, and transaction reliability. Let the equipment operation data feature vector be... The feature vector of electricity consumption behavior pattern data is The feature vector of power grid interaction data is The feature vector of market-related data is Integrate them into a comprehensive feature vector The calculation formula is: ,in, Weighting coefficients are pre-defined based on the importance of the data, and , This indicates a feature fusion operation, specifically the addition of corresponding elements, which is then encrypted and stored in the blockchain distributed ledger, such as... Figure 1 As shown.

[0054] S200, Credit Modeling and Initial Assessment: An initial credit scoring model is created based on a fuzzy membership function, and the initial credit scoring model is trained and optimized using a comprehensive feature vector. The initial credit score of the transaction entity is calculated through the optimized model and recorded on the blockchain.

[0055] Specifically, S200 includes:

[0056] Model Construction and Training: An initial credit scoring model is created using a fuzzy membership function model, where the fuzzy membership function is denoted as . This is used to measure the impact of data features on credit scores, with the initial credit score being... The calculation formula is: ,in, For the comprehensive feature vector The first in One element, for The fuzzy membership degree belonging to high-credit characteristics is determined by a fuzzy rule table. It is the feature vector of equipment operation data. The number of elements, It is a feature vector of electricity consumption behavior pattern data. The number of elements, It is the feature vector of power grid interaction data. The number of elements, It is a feature vector of market-related data. The number of elements is determined based on power grid industry standards. Pre-defined rules govern the impact of different data characteristics on credit scoring (e.g., "harmonic exceedance" reduces credit membership, "on-time settlement" increases membership). The user's historical comprehensive feature vector (e.g., equipment operation data and power quality records from the past 12 months) is divided into training and testing sets in a 7:3 ratio. The fuzzy membership function parameters are adjusted through iterative training. The F1 score is used to verify the model's generalization ability. The formula for calculating the F1 score is: Where TP is a true positive example, representing a correctly predicted positive sample; FP is a false positive example, representing a incorrectly predicted positive sample; and FN is a false negative example, representing an incorrectly predicted negative sample. The model is iteratively trained until the F1 score is ≥0.85. The final fuzzy rule parameters and model structure are then solidified and deployed to the credit assessment module of the blockchain system for the initial credit score calculation of the transaction entity, ensuring that the score results accurately reflect the user's true credit level.

[0057] The user's initial comprehensive feature vector (including historical harmonic exceedance records) is input into the initial credit scoring model, and the initial credit score is calculated using a fuzzy membership function model. The equipment scored highly in operational stability, but due to past violations of power quality standards, its overall credit membership was low, resulting in an initial credit score of [insert score here]. The results are then synchronized to the blockchain credit profile.

[0058] Abnormal event trigger monitoring: At 10:00 AM on a weekday, the system detects through sensors that the output current of the user's photovoltaic inverter drops sharply by 50% (below the normal power generation threshold), and at the same time, the grid side detects that its grid-connected harmonic content exceeds the standard by 15% (exceeding the national standard limit). The system determines that the equipment is abnormal and the power quality is unqualified, triggering the credit score update process.

[0059] S300, Real-time Monitoring and Scoring Update: Real-time tracking of the operating status of power generation equipment and user electricity consumption behavior. Once equipment failure or abnormal electricity consumption is detected, the latest comprehensive feature vector in the blockchain is retrieved, combined with the credit score and feature change amount at the previous moment, and the real-time credit score of the transaction subject is updated through the real-time credit scoring algorithm and synchronized to the blockchain.

[0060] Specifically, the S300 includes:

[0061] Feature data retrieval and analysis: Retrieve the user's latest comprehensive feature vector from the blockchain and combine it with the credit score from the previous moment. The real-time credit score is updated through a real-time credit scoring algorithm. Let the real-time credit score be... The calculation formula is: ,in, This is the score from the previous moment. For the period from the last rating to the current moment The change in each characteristic This is a time decay factor that decreases as the scoring time interval increases. The coefficients are dynamically adjusted based on the degree of data anomalies. It is the feature vector of equipment operation data. The number of elements, It is a feature vector of electricity consumption behavior pattern data. The number of elements, It is the feature vector of power grid interaction data. The number of elements, It is a feature vector of market-related data. The number of elements, a sudden drop in equipment current reflects a decrease in power generation stability, corresponding to a negative change in characteristic value, and excessive harmonics are a highly sensitive risk characteristic, which significantly reduces credit membership.

[0062] Based on the output of the real-time credit scoring algorithm, the user's credit score is calculated from... Downgraded to The updated rating information is immediately synchronized to the blockchain, allowing power grid companies, regulatory authorities, and users to access it.

[0063] S400, Risk Modeling and Quantitative Assessment: Analyzes market data stored on the blockchain, and based on the analysis results, constructs a risk assessment model using the real-time credit scores of the trading entities and market price fluctuation data as inputs, and calculates the risk value of each electricity transaction;

[0064] Specifically, the S400 includes:

[0065] Risk Value Calculation: The risk assessment module, based on a risk assessment model, uses the user's real-time credit score and market data (such as current fluctuations in photovoltaic feed-in tariffs) as input to calculate the transaction risk value. The calculation formula is as follows: ,in, The average credit score in the market. These are the weighting coefficients, and , Market data characteristics Risk sensitivity coefficient Current credit score It is a feature vector of market-related data. The number of elements, credit score below the market average, triggers credit risk weighting items, harmonic exceedance corresponds to a high risk sensitivity coefficient in the past, amplifying market data risk items.

[0066] S500, Risk Warning and Smart Trigger: Based on market conditions, a risk warning threshold is preset. When the trading risk value reaches the threshold, the warning trigger probability is calculated. When the probability exceeds the preset threshold, the smart contract is automatically triggered to send warning information to relevant trading parties and regulatory agencies.

[0067] Specifically, the S500 includes:

[0068] Warning Triggering and Handling: When the risk value reaches the preset risk warning threshold... At that time, the system assesses the urgency of the risk through an early warning trigger probability calculation model. The calculation formula is as follows: ,in, The parameter is used to adjust the sensitivity to probability changes, and its value ranges from 0.5 to 2.0. If the calculated probability of triggering the warning exceeds the threshold of 0.7, the smart contract will be automatically triggered.

[0069] The smart contract executes the following: It sends SMS notifications to users, requiring them to investigate inverter faults and address harmonic issues within 24 hours; it notifies the power grid maintenance team to conduct on-site inspections the following day; and it adjusts the user's grid-connected power limit (e.g., from full grid connection to a 50% power limit) until the rectification meets the standards. The entire process of warnings and handling is recorded on the blockchain to ensure traceability.

[0070] In summary, in long-term power transactions between power generation companies and industrial users, this invention integrates equipment operation, electricity consumption behavior, grid interaction, and market credit data through multi-dimensional data collection and fusion. It constructs an initial credit score using a fuzzy membership function model and dynamically updates the score through a real-time credit scoring algorithm. The risk assessment module combines the risk assessment model with an early warning trigger probability calculation model to achieve automatic risk identification and smart contract triggering. Figure 2 As shown, this mechanism relies on blockchain to ensure data immutability, dynamically responds to equipment failures and market fluctuations, effectively prevents default risks, enhances the trust foundation and regulatory efficiency of large-scale transactions, and is suitable for high-value, long-cycle electricity market transaction scenarios.

[0071] Example 2:

[0072] Please see Figure 2 , Figure 2 The flowchart illustrates the system construction for a blockchain-based trust mechanism in the electricity market, used for long-term trust assessment of electricity transactions between power generation companies and industrial users.

[0073] Data Acquisition and Fusion Module: This module collects equipment operation data, electricity consumption behavior data, grid interaction data, market transaction data, and grid operation status data through sensors deployed on power equipment, smart meters, grid-side monitoring systems, and power trading platforms. After preprocessing, the data is integrated into a comprehensive feature vector using a multi-dimensional feature fusion algorithm and then encrypted and stored in a blockchain distributed ledger.

[0074] The credit modeling and initial assessment module constructs an initial credit scoring model based on fuzzy membership functions. It trains the model using historical data and optimizes parameters to calculate the initial credit score of the transaction entity. During model training, training and testing sets are divided, and the generalization ability is verified using F1 scores. Once the model meets the requirements, it is deployed to the blockchain.

[0075] Real-time monitoring and scoring update module: Tracks device operating status and user behavior in real time. When an anomaly is detected, it calls the latest data in the blockchain, combines the time decay factor and dynamic adjustment coefficient, updates the credit score through the real-time credit scoring algorithm, and synchronizes it to the blockchain.

[0076] Risk Modeling and Quantitative Assessment Module: Using real-time credit scores and market data as input, this module constructs a risk assessment model and calculates the transaction risk value.

[0077] Risk warning and intelligent triggering module: preset risk threshold. When the risk value exceeds the threshold, the warning probability is calculated through a logic function, and the smart contract is triggered to execute the warning notification and power adjustment operation. The entire process is traceable by recording data on the blockchain.

[0078] Taking this scenario as an example, in the provincial electricity market, large power generation companies (such as coal-fired power plants) sign annual electricity trading contracts with high-energy-consuming industrial users (such as steel mills), involving huge amounts of electricity transactions. Both parties need to dynamically assess the credit risk of the transaction, prevent the risk of default, and at the same time meet the requirements of power grid supervision.

[0079] Data acquisition and fusion module:

[0080] Equipment operation data collection: By deploying temperature sensors, vibration sensors, and current sensors in the generator sets of power plants, key operating parameters of the equipment (such as bearing temperature, vibration frequency, and current intensity) are collected in real time. The built-in monitoring module obtains the equipment running time, number of start-stop cycles, and fault codes. Similar operating data is collected synchronously from the user's production equipment (such as blast furnaces and rolling mills).

[0081] Electricity consumption behavior data collection: Utilize smart meters to continuously monitor users' real-time active power, reactive power, and power factor, and analyze their behavior patterns during peak electricity consumption periods and load stability.

[0082] Grid interaction data collection: Obtain voltage stability indicators (such as voltage deviation range) during power transmission from the grid-side monitoring system, and record power interaction data (such as injected or consumed power values) between users and the grid.

[0083] Market and credit data collection: Access the power trading platform to obtain time-of-use transaction prices (such as peak-valley electricity prices) in real time, and collect historical transaction records and default information of both parties (such as whether settlement was made on time and whether dispatch instructions were violated).

[0084] Data processing: The collected data is preprocessed and features are extracted. A multi-dimensional feature fusion algorithm is used to fuse the features and generate a comprehensive feature vector. The formula is as follows: It is encrypted and stored in the blockchain distributed ledger.

[0085] The initial credit score model is constructed using a fuzzy membership function, and the calculation formula is as follows: Based on historical preset model parameters (such as the weights of each data feature on credit), the historical comprehensive feature vector is proportionally divided into training and testing sets. Model parameters are adjusted through iterative training, and the F1 score is calculated using the testing set to verify the model's generalization ability. The calculation formula is as follows: To ensure scoring accuracy, if the F1 score is ≥0.85, the model meets the standard and is considered to have sufficient generalization ability, and training is stopped. If it does not meet the standard, the training and verification process is repeated until the indicator requirements are met. After verification, the final fuzzy rule parameters and model structure are solidified and deployed to the credit assessment module of the blockchain system to calculate the initial credit scores of power generation companies and users and record them on the blockchain.

[0086] Credit Modeling and Initial Assessment Module: The system continuously tracks the status of generator sets of power generation companies and the electricity consumption behavior of users. On weekday mornings, if the vibration sensor of the power generation company detects abnormal vibration of the steam turbine (triggering an equipment fault warning), the system immediately retrieves the latest comprehensive feature vector from the blockchain and calculates the impact of feature changes on credit through a real-time credit scoring algorithm. The calculation formula is as follows: .

[0087] Considering that equipment failures could lead to a decrease in power supply reliability, the credit score of power generation companies has been adjusted from... Temporarily downgraded to This information is simultaneously updated to the blockchain for query by trading parties and regulatory agencies.

[0088] Risk Modeling and Quantitative Assessment Module: Based on the risk assessment model, and combining real-time credit scores of both parties with market price fluctuation data (such as recent electricity price increases), the transaction risk value is calculated using the following formula: Due to the downgrade of power generation companies' scores, the risk value is approaching the preset threshold. .

[0089] The system further calculates the probability of triggering the warning, using the following formula: When the probability exceeds the threshold of 0.7, the smart contract is automatically triggered to send a risk warning to the user and simultaneously notify the power grid dispatch center to strengthen power supply monitoring and prevent the risk of transaction interruption.

[0090] In summary, this embodiment, based on blockchain technology, constructs a trust mechanism for electricity market transactions. Focusing on the annual electricity trading scenario between power generation companies and high-energy-consuming industrial users, it integrates multi-source data—including equipment operation, electricity consumption behavior, grid interaction, market data, and credit data—through a data acquisition and fusion module. After preprocessing and feature fusion, the data is encrypted and stored on the blockchain to ensure its authenticity and reliability. The initial credit scoring model uses a fuzzy membership function to train the model, and F1 scores are used to verify its generalization ability, determining the initial credit score for the trading entity. The real-time monitoring and scoring update module uses time decay factors and dynamic adjustment coefficients to update the credit score in real time based on equipment anomalies and other factors. The risk modeling and quantitative assessment module combines the credit score and market data to calculate the transaction risk value. The risk warning and intelligent triggering module uses preset thresholds and logical functions to trigger smart contracts to execute warning and control operations when the risk exceeds the limit. The entire system achieves dynamic assessment and prevention of credit risk throughout the entire electricity trading process, protecting the rights and interests of both trading parties while meeting grid regulatory requirements, effectively reducing default risk, and providing technical support for the stable and reliable operation of the long-term electricity market.

[0091] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1.A method for constructing a blockchain-based trust mechanism for electricity market transactions, characterized in that, The specific steps of the construction method are: S100, data acquisition and feature fusion: collect power user equipment operation data, power consumption behavior mode data, power grid interaction data, and power market transaction price fluctuation data, historical default records, power grid operation state data, preprocess and extract features, and fuse into a comprehensive feature vector, and store in a blockchain distributed ledger; S200, credit modeling and initial assessment: create an initial credit scoring model based on a fuzzy membership function, and train and optimize the initial credit scoring model using the comprehensive feature vector, calculate the initial credit score of the transaction subject through the optimized model, and record it on the blockchain; S300, Real-time Monitoring and Scoring Update: Real-time tracking of power generation equipment operation status and user electricity consumption behavior. Upon detecting equipment failure or abnormal electricity consumption, the latest comprehensive feature vector from the blockchain is retrieved. Combined with the previous credit score and feature changes, the real-time credit score of the transaction entity is updated using a real-time credit scoring algorithm. Let the real-time credit score be... The calculation formula is: ,in, This is the score from the previous moment. For the period from the last rating to the current moment, The change in each characteristic This is a time decay factor that decreases as the scoring time interval increases. The coefficients are dynamically adjusted based on the degree of data anomalies. It is the feature vector of equipment operation data. The number of elements, It is a feature vector of electricity consumption behavior pattern data. The number of elements, It is the feature vector of power grid interaction data. The number of elements, It is a feature vector of market-related data. The number of elements is recorded and synchronized to the blockchain; S400, risk modeling and quantitative assessment: analyze the market data stored in the blockchain, input the real-time credit score of the transaction subject and the market price fluctuation data according to the analysis result, build a risk assessment model, set the risk value as , and the calculation formula is: , wherein, is the market average credit score, is the weight coefficient, and , is the risk sensitive coefficient of the market data characteristics , is the current credit score, is the element number of the market related data characteristic vector , and the risk value of each power transaction is calculated; S500, risk warning and intelligent triggering: preset the risk warning threshold according to the market situation, calculate the warning trigger probability when the transaction risk value reaches the threshold, and automatically trigger the smart contract when the probability exceeds the preset threshold, and send warning information to relevant transaction parties and regulatory agencies. 2.The method of claim 1, wherein, The collection method and content of each type of data in S100, data acquisition and feature fusion are: Power user equipment operation data: by deploying temperature sensors, vibration sensors, and current sensors on power user equipment, real-time collection of equipment temperature, vibration frequency, and current intensity operation parameters; through the built-in monitoring module of the equipment, obtain the data of equipment running time, start-stop times, and fault code state; Power consumption behavior mode data: use smart meters to collect real-time active power, reactive power, and power factor data of users; Power grid interaction data: obtain voltage stability data during power transmission from the power grid side monitoring system, and record power interaction data between the power grid and the user; Power market transaction price fluctuation data: access the power market trading platform to obtain real-time power transaction prices at different trading times; Collect historical transaction price data, and calculate price fluctuation range, fluctuation frequency, and price change trend data; Collect historical default records: retrieve the default history of the transaction subject from the power market transaction management system, and record the data of default time, default party name, default type, and default processing result; Power grid operation state data: obtain the overall load rate, line transmission power limit, and transformer load rate of the power grid; monitor the power quality data of the power grid frequency deviation, harmonic content, and three-phase unbalance degree, and the opening and closing state information of the power grid equipment. 3.The method of claim 1, wherein, The S100, the multi-source data is fused into a comprehensive feature vector through a multi-dimensional feature fusion algorithm in data acquisition and feature fusion, assuming that an equipment operation data feature vector is , a power consumption behavior mode data feature vector is , a power grid interaction data feature vector is , and a market related data feature vector is The comprehensive feature vector is integrated as The calculation formula is: Wherein, is a weight coefficient preset according to data importance, and , represents a feature fusion operation, specifically, corresponding elements are added. 4.The method of claim 3, wherein, In step S200, the initial credit scoring model is created by combining fuzzy membership functions with the credit modeling and initial assessment. Let the fuzzy membership function be... This is used to measure the impact of data features on credit scores, with the initial credit score being... The calculation formula is: ,in, For comprehensive feature vectors The first in One element, for The fuzzy membership degree belonging to high-credit characteristics is determined by a fuzzy rule table. It is the feature vector of equipment operation data. The number of elements, It is a feature vector of electricity consumption behavior pattern data. The number of elements, It is the feature vector of power grid interaction data. The number of elements, It is a feature vector of market-related data. The number of elements. 5.The method of claim 1, wherein, The training steps of the initial credit scoring model in S200, credit modeling and initial assessment are: (1) Collect historical comprehensive feature vectors and divide them into training set and test set according to the ratio of 7:3; (2) Pre-set the initial parameters of the fuzzy membership function based on domain knowledge, and define the fuzzy rule table; (3) Input the training set into the initial credit scoring model, iteratively adjust the parameters based on the training set, and calculate the F1 score using the test set to verify the generalization ability of the model, and the calculation formula is: Wherein: TP is the true positive, indicating the positive sample predicted correctly, FP is the false positive, indicating the positive sample predicted incorrectly, FN is the false negative, indicating the negative sample predicted incorrectly; (4) When F1 score ≥ 0.85, the model meets the standard, and the model generalization ability meets the standard, stop training; if not, return to step (3), adjust the fuzzy rule parameters and repeat the training and verification process until the index requirements are met; (5) After verification, the final determined fuzzy rule parameters and model structure are solidified and deployed to the credit evaluation module of the blockchain system for initial credit score calculation of the transaction subject. 6.The method of claim 1, wherein, The S500, the calculation of the early warning trigger probability in the risk early warning and intelligent triggering, sets the risk early warning threshold as When , the early warning mechanism is triggered, and the early warning trigger probability is calculated, which is set as W, and the calculation formula is: , wherein is an adjustment parameter, used to control the sensitivity of the probability change, and the value range is 0.5-2.0, and when the probability threshold 0.7 is exceeded, the early warning is triggered.

Images