Research on Risk Assessment Method of Enterprise's Arrears Based on Block-Chain Credit Consensus

Background:At present, the power supply market has always acquiesced to the rule of ‘use electricity rst, pay later’,some electricity users may delay payment or even default on electricity bills due to various reasons, causing problems such as a long recovery period for electricity bills and diculty in debt settlement. In order to reduce this kind of phenomenon, the power supply company must understand the user's historical electricity consumption data, capital status, credit status and other information, but the establishment of such a database requires a lot of time and human resource costs. Methods:Based on the distributed storage technology of blockchain, this paper abstracts power supply companies, power users, banking nancial institutions, government regulatory agencies, etc. into nodes on the alliance chain.After that,this paper introduces a credit scoring model to judge the credit rating of user information based on characteristic indicators, and select the corresponding electricity fee recovery policy after the result is obtained, so as to reduce the operating risk of the power supply company. Results:This article combines the power and energy market with blockchain technology to establish a secure and distributed credit data blockchain, and at the same time establish a credit scoring model based on expert review questionnaire data. The analysis results show that this mechanism is suitable for credit data storage and sharing in energy transactions. Conclusions:Research and analysis indicate that the credit risk assessment method of electricity transaction data proposed in this article provides a theoretical basis for the combination of electricity transaction credit risk assessment and blockchain technology,which will help improve the company's ability to assess the risk of arrears and reduce the operating risk of power supply companies.


Introduction
In the current market competition, companies are under greater operating pressure, and many companies are facing closure and bankruptcy. For a long time, power supply companies have been adopting the market rule of "use electricity rst, pay later", so there are hidden dangers such as long recovery period of electricity bills and di culty in reminders. Due to the existence of high-risk consumers, it is very di cult to collect electricity bills, which has a huge impact on the normal operation and development of the company [1]. Once customers have repeatedly delayed or even failed to pay their electricity bills for some reason, and local administrative intervention, short-term receivables of power companies will become mid-term receivables, long-term receivables, and even bad debts. This type of operation of consuming rst and paying bills inherently has certain operating risks. Some customers owe large amounts of electricity bills during the operation of the power system, which seriously hinders the healthy development of the power industry [2]. Based on all responsibility to customers, collection of electricity bills has always been a major problem that plagued power supply companies. Coupled with the impact of the market economy, some poorly managed companies went bankrupt, and some unscrupulous companies passed their own losses to the power supply company in order to reduce losses, and maliciously defaulted on electricity bills, which made it more di cult to recover electricity bills. However, the power supply company lacks effective penalties for such problems and can only restrict it by cutting off the power and charging a small amount of liquidated damages. These methods are low in binding force, leading to frequent arrears of electricity bills, affecting the daily operation and development of power supply companies, which makes the recovery of electricity bills increasingly a major problem that plagues power companies. In order to avoid this situation as much as possible, it is necessary for power supply companies to understand the historical power usage information data, capital status, and credit status of the company before power transactions to formulate a reasonable electricity fee recovery policy and conclude a complete power supply contract. Use this method to determine a reasonable electricity bill settlement method, do a good job of credit analysis, and actively master information and business market conditions.
Since the power company cannot accurately discover the user's arrears, resulting in a shortage of power transaction data [3]. This enables power supply companies to spend a lot of manpower and material resources to store and manage corporate power consumption data while completing normal production and operation, which greatly increases the burden on power supply companies. The electricity consumption information of an enterprise is complex and diverse, and its storage requires a special database and management by professionals. But this does not guarantee that power supply companies can easily and accurately obtain the information they need. The fundamental reason is that there are many problems with centralized storage.1. Higher cost. 2. The data transmission speed is slow. 3. Low security. 4. Privacy leakage.
In the current power system, distributed energy makes it possible for high-consumption customers to signi cantly reduce the share of the remaining costs they pay, leaving the de cit to the power supply company [4]. In order to reduce the operating pressure of power companies, various power companies have proposed various administrative and technical means, trying to establish user-oriented power recovery risk assessment strategies. However, due to the insu cient ability of power grid companies to predict the risk of arrears of power users, they cannot take targeted preventive measures in advance according to the risk level of customer arrears, which is a major problem facing the prevention of the risk of electricity fee recovery. In order to effectively improve the risk prevention and control capabilities of power companies and reduce business risks, e cient and accurate arrears risk prediction is particularly important.
With the development of information technology in the new era, distributed storage technology based on blockchain can effectively liberate the above problems. Distributed storage of data is not stored on a single company's server, but is encrypted and distributed through a decentralized network. It has higher reliability, availability and lower cost. This paper uses the DPOS consensus system in the blockchain to establish a distributed and decentralized credit data processing method. The information database is jointly maintained by all members of the chain, and the cost is low. All enterprise electricity information data is encrypted and stored on the blockchain, which can not only ensure the privacy of the enterprise, but also realize that each transaction of the enterprise account can be traced on the chain for easy query.This blockchainbased decentralized structure avoids the possibility of a node tampering with information privately and increases the security of the system. In this way, the power supply company can inquire about the basic data, electricity bill information, historical payment records and other data of the electricity user, and can conduct risk assessment and analysis on the user's historical data. As a result, a credit evaluation method for power users is established, user risk levels are re ned, and exible electricity bill settlement methods are adopted according to user arrears risk levels to optimize the electricity bill recovery process and reduce business risks.

Literature Review
For a long time, due to its traditional monopoly, Chinese power industry has faced increasing pressure from peer competition and technological shocks with the advancement of the market system. As the last link in the management of power supply enterprises, electricity fee recovery management has always been the focus of power supply enterprises [6], and effective electricity fee recovery management is inseparable from effective data management and data analysis.
With the development of economy and technology, data mining and big data analysis techniques are often used in China to analyze users' electricity consumption characteristics, so as to nd out the relationship between electricity consumption characteristics and the risk of arrears. Li Jianbin et al. [5] used the big data analysis method to extract and transform data, and established the risk identi cation model of customer electricity bill by machine learning. While performing big data mining on user data, Zhao Yadi et al. [6] introduced logistic regression and scoring card function methods to screen and transform index data. Zhao Yongliang et al. [7] also took big data analysis and logistic regression model as the design connotation, and proved the validity of the risk prediction model through example analysis.
[8] formed a user payment data release model based on data mining and analysis of user payment methods, payment frequency, payment time, electricity consumption changes, and electricity business changes to carry out electricity tariff risk level assessment.
In addition to big data analysis, Xie Lefeng et al. [9], facing the risk of electricity charge recovery, proposed to make early warning of customers' electricity charge recovery risk based on the long and short term memory network algorithm in deep learning, and used entropy method and correlation to analyze key indicators of electricity charge recovery risk, and achieved remarkable results. Wu Haiguang et al. [15] proposed a comprehensive learning method --gradient enhanced decision tree --random forest --Adaboost logistic regression (GRAL), and established a risk prediction model for electricity price recovery. Experiments show that this method can effectively evaluate the payment behavior of power users.
Foreign countries have also studied the issue of electricity fee recovery. Tim Schittekatte's team [11] studied the problem of electricity charge recovery from the perspective of electricity price design, described the problem of power grid cost recovery of distribution system as a non-cooperative game between consumers, and focused on the limitation of capacity-based charging, which will provide challenges for power grid cost recovery. Batlle et al. [12] also studied cost recovery from the perspective of residual cost distribution of electricity price. They analyzed the problem according to the rate setting theory, reviewed the different allocation methods currently considered, and proposed a solution based on unbalanced xed fees to achieve e ciency, fairness and cost recovery of modern electricity price.
However, the problem of electricity bill collection does not stop there. More importantly, it is based on the judgment of user credit. Abdulrahman [13] and others took KWIQPLUS loans as an example, using the fuzzy logic method of Ghana's micro nance credit scoring to complete the evaluation of user credit.
Based on Turkish credit card data, Akkoc and Soner [14] compared traditional technology, neural network, and three-stage hybrid adaptive neuro-fuzzy inference system (ANFIS) credit scoring analysis model,and showed that the model is superior to traditional data analysis methods in terms of the average correct classi cation rate and the estimated cost of misclassi cation, which is a good reference for the credit evaluation of users in the power industry.
So far this paper have come to the conclusion that in order to effectively establish user risk assessment models, reasonable data analysis and data sharing technologies are indispensable.With the advent of the information age, while big data reduces the cost of credit, it also brings about the problems of big data monopoly and misplaced information ownership [10]. Facing the problems of information asymmetry and lack of credit between power supply companies and users, the current research methods are mostly based on data storage and data sharing based on blockchain technology. In addition, the blockchain credit consensus mechanism can better solve the credit problem between power supply companies and power users. In the existing domestic literature, there are many articles on the blockchain consensus model. Yang Man [16] introduced blockchain into the existing Internet group collaboration, and constructed a group collaboration model based on blockchain trust consensus technology, which can be widely used in enterprises. In power supply companies, it is precisely due to the lack of mutual trust and consensus mechanism that transaction information is not shared, which leads to the risk of recovery of power supply companies. In the current Internet environment of information asymmetry,Li Shuai [17]created a new operation mode through the innovative information consensus mechanism and trust mechanism of block chain. The research focus is on how to transmit and maintain information data in the chain, which has certain enlighten effect on the research of this paper. The management and sharing of information and data in the construction eld by the blockchain credit consensus mechanism is also related to the power industry.Taking this as the research object, Li Meng et al. [18] used the unique function of block recording data to trade building materials into a complete information trading chain.Cao Yang et al. [19] built a construction supply chain information sharing management blockchain to provide a new model for information sharing in the construction supply chain to promote information sharing among project participants in the supply chain.
In foreign countries, the research on blockchain and reputation privacy is more extensive. Keke Gai et al.
[20] proposed a block chain edge model suitable for smart grid in order to protect user privacy of smart grid, and ensured the security of energy transactions by using smart contracts. Nir Kshetri [21] emphasized the superiority of blockchain for user privacy and transaction security in the eld of Internet of Things. Also in the eld of the Internet of Things, Seong-Kyu Kim et al. [22] proposed a blockchainbased ecosystem to improve e ciency and stability in view of the multi-platform security vulnerabilities of the Internet of Things. Seong-Kyu Kim and Jun-Ho Huh[24] used block chain technology to measure carbon emission rights and proposed a carbon emission right veri cation system based on block chain. It also discusses the veri cation encryption algorithm required by the blockchain engine.
This idea is also applicable to the power industry. At present, distributed energy is a global hot issue. Wayes Tushar et al.
[23] tried to study P2P distributed energy networks through game theory, and used auction theory models to simulate the actions of participants in the grid for effective energy management. Italo Atzeni et al. [25] proposed a new algorithm of distributed intelligent instrument, which not only protects privacy, but also provides an optimal production and storage strategy, and is directly applicable to small enterprises and home end users.
Nowadays, the combination of blockchain technology and distributed energy trading is the general trend. Merlinda Anonia and Valentin Robu[26] studied more than 140 block chain research projects and startups based on the system architecture of block chain technology, and concluded the potential and feasibility of building block chain for energy application. Among them, the blockchain consensus issue is the top priority.In order to study the credit consensus mechanism of blockchain, Yuhao Wang et al. studied the dynamic consensus mechanism based on reputation and applied it to real-time energy transactions to solve problems such as low transaction throughput and long delay time.
In addition, in terms of data processing and data sharing, blockchain technology can also be used to great advantage.Jiawen Kang et al.
[31] put forward a reputation-based data sharing scheme in the vehicle edge network system by using Alliance blockchain and smart contract technology to ensure highquality data sharing of vehicles. Jian Shen et al.
[32] also studied data sharing, proposed to provide high security and high e ciency data sharing and data storage for the same group in the cloud in an anonymous way, and established a new data sharing scheme of traceable group to support anonymous multiple users in the public cloud.
From the above point of view, blockchain can be applied to a variety of social elds due to its characteristics of decentralization, immutability of information, openness and independence, which will provide a good reference for it in the eld of power trading.In this paper, based on the alliance chain technology in the blockchain, the nodes jointly record the transaction credit data through the on-chain transaction, and pack it into blocks through the encryption algorithm, so as to achieve the data's immutability and traceability.
The main contributions of this paper are as follows: 1. This paper proposes to build a secure and distributed credit data blockchain using the Alliance Blockchain technology for credit data storage and data sharing in energy transactions; 2. In the blockchain, the DPOS consensus mechanism is used to select authorized nodes to achieve e cient storage of energy data and data sharing between the two parties in the transaction, and at the same time protect the privacy of users with an asymmetric encryption algorithm; 3. This paper deploys energy transaction smart contracts in the blockchain, which can realize the automatic execution of the transaction when they are determined, and the transaction data will also be packaged on the chain; 4. Analytic Hierarchy Process (AHP) model is introduced to evaluate user credit rating at volunteer nodes to help power supply companies choose appropriate power recovery policies.
The rest of this article is arranged as follows. The rst part of the third section introduces the application architecture of alliance chain based on P2P technology. The second part of the third section describes the steps of information and data storage and data sharing for each component node of the transaction blockchain. The third part introduces the credit evaluation model, the model uses the analysis method to screen the sample indicators, and then uses the analytic hierarchy process to calculate the weight of the model, so as to score the information provided by the user and judge the user's credit rating and choose the appropriate electricity fee recovery method. The fourth section summarizes the conclusion of this paper.

Methods
In recent years, blockchain technology has developed rapidly. It is known from the above-mentioned literature that it is widely used in various industries. In this paper, blockchain technology is introduced into the eld of power transactions, and the value of blockchain is extended to a wide range of services ranging from pure data storage (such as device con guration and governance, sensor data storage and management, and multi-access payment) [27]. In the eld of power recycling transactions, this paper studies new transaction models based on blockchain credit consensus, and builds a distributed information storage method based on digital encryption and veri cation technology to provide power users with non-tampering and traceability for historical credit data. At the same time, this paper designs the application of user credit query method based on alliance blockchain, and elaborates on application architecture, information storage, credit query logic and so on.Firstly, combining with the technical characteristics of blockchain, the storage and query mechanism of information on the chain is described.
Finally, according to the user's information on the chain, the risk assessment and analysis are carried out to determine the risk level, so as to implement the corresponding electricity bill settlement method.
1. Consortium chain application architecture Blockchain technology was proposed by Satoshi Nakamoto in 2008. After continuous iterative evolution, it has formed ve major technical features: distributed, trustless, time stamping, asymmetric encryption and smart contracts. Compared with the operating space and e ciency of the private chain, the value of the alliance chain is greater. Compared with the uncontrollable and privacy security issues of the complete decentralization of the public chain, the alliance chain is more exible and more maneuverable.According to the transaction characteristics of electricity companies, power supply companies and nancial institutions, this paper abstracts the regulatory agencies such as companies and nancial institutions as nodes on the alliance chain for research.
Blockchain data is timestamped, so it can be used as a reliable tool for data traceability of power and energy transactions. Based on blockchain technology, this paper establishes a distributed and decentralized credit data storage mechanism, and obtains the consensus of banks and nancial institutions, government regulatory agencies, large power supply and sales companies, and power users, and forms a joint management by the three parties. The blockchain data architecture is mainly divided into six levels: data layer, network layer, consensus layer, incentive layer, contract layer, and application layer, which jointly complete transaction veri cation, judgment, storage, and backup [9].

Risk assessment of arrears based on alliance chain technology
Alliance chain is a relatively new way to apply blockchain to enterprises, which can be applied to collaborative services between various enterprises. In the alliance, a number of nodes are pre-designated for accounting, and the nodes that join later need to apply and verify, and join the alliance chain after approval. This paper introduces volunteer nodes [19] as the lower-level virtual members of power supply enterprise nodes in the blockchain to help power supply companies complete data analysis and risk assessment.
The form of alliance chain is used for the traceability and sharing of power consumption data information of power companies, which can ensure that all parties to the transaction share information with each other. At the same time, the non-fully decentralized structure guarantees the relative leadership of power supply companies, bank governments and other regulatory agencies in the entire organization, and the transaction speed of the alliance chain can also handle various transaction information data generated during the transaction process[20 ].
(1) Division and role of institutions on the alliance chain In the entire alliance node, the institutions and functions are as follows: Power supply enterprise node Containing institutions a supply node that receives transaction requests and provides electricity and energy services.
Information data The information provided includes power energy quality, stock quantity, energy production time, power plant location, energy transportation routes, power energy purchase and transportation price and other basic energy information.

User Enterprise Node
Containing institutions The demand-side unit composition of electric energy includes high-voltage nonresident users and low-voltage non-resident users.
Information data The information provided includes basic information such as the number of electricity transactions, expected prices, and historical payment records.
Bank nancial node Containing institutions The composition of banking institutions undertaking corresponding nancial transfer tasks.
Information data The information provided includes basic information such as user company registered capital, nancial loan status, other energy transaction status (such as water charges), historical electricity payment records, etc.

Government regulator node
Containing institutions Unit nodes responsible for power energy operations and trading tasks.
Information data Provides energy policy updates and access to information about the energy trading process.

Volunteer node
Containing institutions The nodes included in the power supply company's subordinates are used to calculate and evaluate the risk level of power users, and assist in completing calculations and transmitting information.
Information data This node is a virtual node, which does not need a secret key. It is only used to receive the information transmitted by the power supply node, perform the corresponding calculation, and then pass the calculation result to the power supply node.
The energy and power information database is jointly maintained and operated by all nodes, and each node has equal rights to exchange and store information. At the same time, data is uploaded through consensus mechanism and P2P network diffusion, and a traceable and tamper-free block is formed by relying on hash algorithm and linked to form a block chain in turn.
(2) Blockchain information storage and sharing According to the different participants in the transaction alliance chain, the alliance node in the blockchain can be divided into banking and nancial node, power supply enterprise node, user node, government regulator, volunteer node and other auxiliary nodes. After each node is veri ed, it will get the corresponding ID IP of its own node.In order to protect the privacy of corporate information and data of power users, this paper adopts asymmetric encryption method, and the public key and private key of each node are generated by the elliptic curve algorithm. Public and private keys appear in pairs, which can not only protect the basic information privacy of electricity users, but also complete the storage and sharing of information and data. After the public and private keys are generated, they broadcast their own information and public keys to the entire network, and obtain and save information from other nodes to form an alliance transaction chain.
Analogous to the blockchain data storage and sharing scheme adopted by Kang in the eld of vehicle interaction [21], in the power information sharing blockchain, the steps of generating and storing the user's original information in the blockchain are as follows: 1: Both parties of power supply and demand, as well as banks and governments, jointly form alliance nodes. The identity of enterprise nodes needs to be veri ed by government regulatory agencies.Each node uses elliptic curve algorithm and asymmetric cryptographic algorithm to obtain public key K and private key k as identity IP identi cation, and generates its own digital signature sig, and at the same time con rms the authority of each node; 2: Each node uses the DPOS consensus mechanism to vote for the authorized node, which is generally voted by all enterprises on the alliance chain. Each unit uploads its own information through the P2P network, which is packaged and uploaded to the block chain by the authorized node to store the original data of information; 3: When the user company needs to purchase energy, the node Xi sends the transaction request to the authorized node with the time stamp and digital signature sig, and the node veri es its identity and then feeds back Xi. After that the node Xi will send the transaction history information data of the most recent year (Such as: current default situation, historical payment information, corporate credit, etc.)to the authorized node with public key Ki , that is

Vi→AN request sig
Message data|Ki|sig|timestamp Block 4: The authorized node broadcasts the ciphertext information M to other ordinary nodes for review and veri cation.After con rming its authenticity, the information will be packaged and added to the block chain for permanent storage.
Credit information mainly includes two parts: 1. Normal performance (average monthly electricity consumption, frequency of electricity purchase, etc.)

Information on breach of contract (number of breaches, amount of breach, delay in payment of penalties, etc.)
Therefore, the encrypted information of historical data veri ed by users is stored in the alliance chain, which has the characteristics of non-tampering and traceability. In order to simplify the storage space, the recording adopts the method of data indexing. The other nodes of the blockchain can only store the hash value of the data.
After the authorized node veri es the information submitted by the user,it sends its transaction request to the power supply node Yj, Yj queries the user's information data based the blockchain before the transaction, and the volunteer node calculates the risk level according to the index.(The credit index evaluation will be explained in detail in 3 Part).The corresponding electricity fee recovery method will be adopted. The P2P data sharing query steps are as follows: 1: The power supply node Yj sends the identity ip and digital signature to the authorized node when receiving the transaction request. After receiving the response, it queries the data block about Xi according to the information index in the blockchain, and sends a data access request to Xi including its own number signature Yj → Xi request sig|timestamp 2: After receiving it, Xi veri es the identity of YJ according to the digital signature. The permission opening authorization is sent to the authorized node, and the private key of the queryer and the public key of the queryer Yj are attached for query.
Xi → AN message authority|ki|Kj|timestamp 3: Upon receiving the information, the authorized node has the right to open query to Yj .Yj queries Xi's historical transaction and nancial record raw data according to the bank node data pool and the information index in the blockchain, and sends it to volunteer nodes for data analysis after veri cation to determine the user's credit rating (detailed in the next section), so that the corresponding electricity fee transaction measures can be implemented under the control of the government supervision node.
1. The architecture consists of user module, access control strategy module, data analysis module,and transaction module.
2. Electricity supply and use companies obtain their own public and private keys through key handling.
3. Electricity companies set access policies, power supply nodes verify their identities, and obtain permissions 4. If the identity of the power supply node matches the access control policy set by the power user, the encrypted shared data can be obtained through the blockchain and the bank data node, and the data will be sent back to the power supply company after being processed by the key center 5. The volunteer nodes subordinate to the power supply company judge the credit rating of the power supply company according to the data sent back. The power company then transacts power with customers through smart contracts under the supervision of government nodes and uploads it to the blockchain.
(4) Analysis of the alliance chain transaction and veri cation process The power transaction alliance chain established in this article is developed around the historical information and data of user companies. It is a tool for the power supply company to verify transactions.
Combined with the blockchain system, it can further increase the ability of tamper-proof, transparent transaction processing and privacy protection[27] .
Among them, the power transaction smart contract is formulated by the nodes in the alliance chain. The power supply company, the bank supervisory agency and the power user respectively make transaction guarantees. This commitment stipulates the rights and obligations of the parties to the transaction, and is generated after both parties' private keys are signed.
The contract is translated into the corresponding code, broadcast to the entire chain via the P2P network, and after veri cation by all parties, it is transmitted to the blockchain by the authorized node to form an energy transaction smart contract.During the execution of the smart contract, the status information always records the energy status of the power supply company. After the transaction is completed, the grid unit, bank and supervisory unit verify the transaction completion data information.After the veri cation is passed, the authorized node on the chain temporarily stores the information in the transaction database and disseminates it to the entire chain. After the entire chain is veri ed, the transaction information can be packaged on the chain (the process is similar to the second part: data storage on the chain ) to ensure the authenticity and traceability of transaction information.
Before the transaction, the electricity user rst veri es the identity through the "registered account" of the alliance chain node.Relying on the alliance chain technology platform, the node initiates a transaction request in the blockchain, stamps a time stamp and digital signature, forms the transaction cipher text through the hash function and publishes it to the entire network. After receiving the transaction request, the authorized node determines the user's identity information and receives the original data uploaded by the user by checking the digital signature in the transaction request. At the same time, the power supply company packages its own identity and credit information and sends it to the corresponding authorized node, and broadcasts it to all chain.and with the help of authorized nodes, query the user's basic data, electricity bill information, historical payment records and other historical data and current nancial conditions, and send them to the entire network for veri cation.
In the transaction process, both parties of the transaction trigger the power and energy transaction smart contract deployed on the chain by setting the transaction quantity, transaction price and other data.The speci c terms of the transaction contract and the incentive reward sharing for authorized nodes are preset in the smart contract. After the smart contract is over, the transaction information is stamped with time stamp and the digital signature of both parties to the transaction is packaged into the block, and the entire transaction is declared complete .
Grid companies can make detailed explanations on the quality and inventory of electric energy and set prices on their own, and they have the right to change the pricing. Through the deployment of smart contracts, users can initiate transactions with power companies and pay a certain fee, which can achieve the purpose of as-used as needed. The grid company will immediately carry out energy transmission, which reduces the risk of energy users in arrears to a certain extent.

Information extraction and risk level assessment
This paper adopts the analytic hierarchy process model to analyze and score the indicators. This model is a mature forecasting method. First, the analysis method is used to construct the indicator system, and then the indicator weight is calculated through the expert scoring system, thereby establishing an evaluation model for power customers in arrears. [24] 1. Use analytical methods to build an index system The matrix form is The matrix satis es the conditions: According to the ratio scale of the judgment matrix, the experts score to determine the mutual importance of different indicators, and the survey opinion form will be statistically analyzed (Among them: M ij is the arithmetic square root of the importance score of the i-th index and the j-th index, s ij is the sum of the scores, and n represents the number of experts) After completing this judgment matrix, perform a random consistency test and calculate the weights The random consistency test formula is: In the analytic hierarchy process, it is generally believed that when CR<0.1, the judgment matrix has acceptable consistency; when the CR is smaller, the judgment matrix is close to complete consistency, otherwise, the degree of deviation is greater.
RI is a random consistency index, this paper directly borrows from Xu Shubo's RI value of the rst 1-8 steps obtained by repeating the RI of 1-15 steps: After the expert opinions are collected, the score matrices of the sub-target layer and each criterion layer are constructed through statistical analysis. Then, the consistency test is carried out, and the weight value is calculated after the consistency test. The weight value is the proportion in the evaluation system.The calculation method is as follows: At the same time, the eig function in MATLAB was used to calculate the value of the matrix, and then the consistency test was carried out. This time, a total of 30 questionnaires and 30 valid questionnaires were issued.
Through investigation and statistics, this paper obtains the weight distribution of the total target and subtargets Table2. Total target level index weight table  According to the table 2 , the CR value of the total target and the sub-target is 0.068, and the result is less than 0.1, which passes the consistency test.
According to the table 3-6 , the CR value of the total target and the sub-target is 0.068, and the result is less than 0.1, which passes the consistency test.
Through statistical analysis of the questionnaire, the weight distribution of the following sub-target items and various criteria is obtained. Table 3.Weight Table of Business Ability Index   Table 4.Management Ability Index Weight Table   Table 5.Financial Strength Index Weight Table   Table 6.Integrity record index weight table  Power supply companies should carry out certain constraints on management, when necessary, to take a special policy of charge recovery of pay rst and then use electricity, and cautiously sign electricity purchase contracts.
In this section, based on the principles of system, science, and strati cation, the arrears of electricity users are divided into three levels , and then different electricity charge management strategies are formulated for the three levels of customers. While actively maintaining users with good risk levels, targeted measures should be taken to recover electricity charges for users with poor risk levels, so as to avoid operational risks of electric power enterprises, and improve the risk prevention ability of electricity charges recovery in time.

Conclusion
As we all know, the primary characteristics of the blockchain are non-tamperability and openness and transparency. Most scenarios require this technology, but not all companies are willing to make all their information open and transparent.The alliance chain is jointly managed by several institutions or enterprises, and each institution or enterprise runs its own node. The transaction data is only allowed to be recorded by the nodes on the chain. Only the data authorized by the node can be broadcast on the chain. This mechanism well solves the problem of data privacy and the security of all parties. Therefore, alliance chain is the most valuable branch of blockchain at present..
Since the beginning, how to carry out effective risk assessment for power users has been an important work at present [26].It is necessary to nd a specialized database and carry out professional management, but it still cannot solve the problem that the power supply company does not have a comprehensive understanding of user information, which is the disadvantage of central management. Distributed storage using blockchain technology can effectively solve the current central storage of high cost, slow data transmission speed, low security, privacy leakage and other issues.
This article abstracts banking nancial institutions, government regulators, power supply enterprises, and users into alliance chain nodes, expounds the application architecture, information storage and processing, and transaction execution logic of the blockchain credit consensus mechanism, and establishes a power transaction alliance chain model,and gives a credit risk assessment method for electricity transaction data, which provides a theoretical basis for the combination of electricity transaction credit risk assessment and blockchain technology, and helps power supply companies quickly obtain basic user information, electricity preferences, credit risks, etc., which can effectively improve the enterprise's ability to assess the risk of arrears, effectively manage user credit, so as to reduce the operating risk of the power supply company.
Shortcomings of current research in this paper: 1. Chinese energy-related blockchain projects are still in the development and exploration stage, many application scenarios have just landed, and there are still many problems that have not been resolved; 2. This article only considers the research of theoretical methods, and gives theoretical solutions to the risk of enterprise arrears. The speci c practical application needs to improve the mechanism to complete; 3. The arrears risk assessment method of power supply enterprises after information data collection still needs to be improved, the indicators can be clearer and more complete, the application of the model needs more samples to ensure the effectiveness of the prediction, and intelligent assessment is needed when necessary. What is certain is that the consortium chain technology has more research value than traditional blockchain in terms of high performance, availability, programmable, and privacy protection. In addition, future research work can consider the issue of access rights management of transaction data. In distributed storage, using blockchain technology to design on-chain data access rights can further ensure the security of data.

Abbreviations
There are no abbreviations in this paper Blockchain data upload process Blockchain transaction process