Power transaction network risk assessment system based on block chain and edge computing

Abstract

The invention provides an electric power transaction network risk assessment system based on a block chain and edge calculation, and relates to the technical field of electric power transaction and network security, and the system comprises a security assessment framework construction module which is used for building a security assessment framework for assessing electric power transaction information based on the block chain and edge calculation; the index system construction module is used for constructing a three-level index system for power transaction network security risk assessment based on the power transaction data; the power transaction data comprises market price data, transaction volume data, contract data, power generation and power consumption data, settlement data and system state data; the risk level evaluation module is used for carrying out risk evaluation on the power transaction network based on a security evaluation framework and a three-level index system, and carrying out dynamic evaluation on the risk level by applying a BiGRU model; and the risk response module is used for performing intelligent risk response according to the dynamic assessment result of the risk level. According to the scheme, the accuracy of risk assessment on the power transaction network security can be improved.

Description

Technical Field

[0001] The present invention relates to the field of power trading and network security technology, and in particular to a power trading network risk assessment system based on blockchain and edge computing. Background Art

[0002] At present, the digital transformation of the energy sector is in full swing. The new power system generally includes power stations, power grid dispatching platforms, non-adjustable loads, load aggregators, demand-side resources, etc. Power stations refer to thermal power generation, which is used to produce power resources. According to the power demand, the power grid dispatching platform transmits power resources to load aggregators and non-adjustable loads. The non-adjustable load side represents load resources whose power demand cannot be adjusted, such as power-sensitive equipment and loads that cannot be interrupted during power consumption. Load aggregators aggregate adjustable load resources in different spaces to achieve monitoring and control, and participate in power system operation and power market transactions. Third-party institutions, power sales companies, electric vehicle charging stations, etc. are all typical load aggregators. Load aggregators integrate the demand side's demand for buying and selling electricity, allocate electricity according to certain strategies, and realize the flexible control of demand side loads, thereby reducing the pressure on traditional power grids. Demand side resources are composed of adjustable loads, including commercial controllable loads, industrial controllable loads, electric vehicles, etc.

[0003] However, with the rapid development of new power systems, the penetration of distributed energy in distribution networks has increased year by year, and the power trading model has gradually changed from centralized to distributed, which has led to problems such as insufficient mutual trust among market players, low data security, and difficult management. Therefore, risk assessment of power trading network security is of great significance to provide security for the overall power system.

[0004] At present, traditional network security risk assessment methods mainly include qualitative, quantitative, and a combination of qualitative and quantitative methods. However, the current assessment methods are generally centralized and static, and the evaluation indicators are too single, resulting in poor accuracy in risk assessment of power trading network security. Summary of the invention

[0005] In view of this, in response to the above shortcomings, it is necessary to propose a power trading network risk assessment system based on blockchain and edge computing to improve the accuracy of risk assessment of power trading network security.

[0006] The present invention provides a power trading network risk assessment system based on blockchain and edge computing, comprising: a security assessment framework building module, an indicator system building module, a risk level assessment module and a risk response module;

[0007] The security assessment framework building module is used to establish a security assessment framework for evaluating power transaction information based on blockchain and edge computing;

[0008] The indicator system construction module is used to construct a three-level indicator system for power trading network security risk assessment based on power trading data; wherein the power trading data includes: market price data, transaction volume data, contract data, power generation and consumption data, settlement data and system status data;

[0009] The risk level assessment module is used to perform risk assessment on the power trading network based on the security assessment framework and the three-level indicator system, and dynamically assess the risk level using a recurrent neural network BiGRU model;

[0010] The risk response module is used to perform intelligent risk response according to the dynamic assessment results of the risk level.

[0011] Preferably, the security assessment framework building module includes: a system requirement determination unit, a node deployment unit, a blockchain platform deployment unit, a smart contract deployment unit, a security configuration unit and a monitoring mechanism configuration unit;

[0012] The system requirement determination unit is used to determine the system's requirements for processing data; wherein the system's requirements for processing data include: requirements for processing speed and throughput of power trading data, requirements for data security, privacy and anti-tampering, and the number of nodes that the distributed network needs to support, network topology and data synchronization requirements;

[0013] The node deployment unit is used to determine edge nodes that have sufficient computing power, storage capacity, and network bandwidth and are compatible with the selected blockchain platform and related software; and to arrange the locations of edge computing nodes according to the distribution of the power trading system;

[0014] The blockchain platform deployment unit is used to determine a blockchain platform that meets the requirements of smart contract support, data security and scalability; and install and configure blockchain nodes on each edge computing device, and set communication protocols, ports and network topology between nodes;

[0015] The smart contract deployment unit is used to define the business logic of the power transaction and compile it into a smart contract code; and deploy the smart contract to the blockchain to ensure that the smart contract is correctly executed on the blockchain node;

[0016] The security configuration unit is used to configure node authentication, data encryption and authority control;

[0017] The monitoring mechanism configuration unit is used to establish a monitoring mechanism to track the operating status of the system in real time.

[0018] Preferably, the indicator system construction module includes: a third-level indicator determination unit, a second-level evaluation indicator weight determination unit and a first-level evaluation indicator weight determination unit;

[0019] The three-level indicator determination unit is used to determine the evaluation indicators of each level constituting the three-level indicator system;

[0020] The second-level evaluation index weight determination unit is used to construct a hesitant fuzzy preference relationship matrix according to the evaluation index, and to modify the hesitant fuzzy preference relationship matrix using the matrix consistency evaluation index; and to calculate the weight of each second-level evaluation index using the modified hesitant fuzzy preference relationship matrix;

[0021] The first-level evaluation index weight determination unit is used to perform weighted summation on the weights of the second-level evaluation indexes of the same category to obtain the first-level evaluation index weights.

[0022] Preferably, the evaluation indicators at each level are shown in the following table:

[0023]

[0024]

[0025] Preferably, the second-level evaluation index weight determination unit includes: a preference relationship matrix construction subunit, a consistency check subunit, an expected relationship matrix conversion subunit, a limit matrix determination subunit and a secondary index weight determination subunit;

[0026] The preference relationship matrix construction subunit is used to construct the following hesitant fuzzy preference relationship matrix according to the indicator network hierarchy:

[0027]

[0028] Among them, h ij (p ij ) is used to represent the expert i's evaluation of the third-level evaluation index p i Relative to the evaluation index p j The preference degree of each, i,j = 1, 2, ..., n, the dimension of H matrix is ​​n × n;

[0029] The consistency check subunit is used to calculate the consistency value CR of the hesitant fuzzy preference relationship matrix by the following calculation formula; and, when CR is less than a preset value, modify the hesitant fuzzy preference relationship matrix until CR meets the requirement:

[0030]

[0031] Among them, λ is the eigenvalue of the hesitant fuzzy preference relationship matrix, RI is the random one-time index, CI is the statistical average, and n is the dimension of the H matrix;

[0032] The expected relationship matrix conversion subunit is used to convert the hesitant fuzzy preference relationship matrix corrected by the consistency check subunit into the expected hesitant fuzzy preference relationship matrix through the following calculation formula:

[0033]

[0034] in, It is used to characterize the expected hesitant fuzzy preference relationship matrix, l is a loop variable, and iterates over each element in the modified hesitant fuzzy preference relationship matrix;

[0035] The limit matrix determination subunit is used to calculate the expected hesitant fuzzy preference relationship matrix using hierarchical analysis software to obtain a weighted supermatrix; and to perform limit calculation on the weighted supermatrix to obtain a limit matrix reflecting the dependency relationship between the elements of the network layer;

[0036] The secondary indicator weight determination subunit is used to obtain the weights of all second-level indicators according to the limit matrix.

[0037] Preferably, the risk level assessment module comprises: a data acquisition unit, a data processing unit, a transaction network risk assessment unit and a risk level assessment unit;

[0038] The data acquisition unit is used to acquire power transaction data; wherein the power transaction data includes system data and external interface data;

[0039] The data processing unit is used to clean the power transaction data to obtain risk assessment data;

[0040] The transaction network risk assessment unit is used to evaluate the risk assessment data using the smart contract configured on the security assessment framework, and calculate the risk score according to the three-level indicator system to obtain risk situation ledger data; and generate risk report data from the risk situation ledger data of power transactions based on the Merkle tree;

[0041] The risk level assessment unit is used to dynamically assess the risk level by applying the recurrent neural network BiGRU model through historical data and current risk report data.

[0042] Preferably, the data processing unit includes a normalization processing subunit and a weighted fusion processing subunit;

[0043] The normalization processing subunit is used to perform normalization processing on the power transaction data by the following calculation formula:

[0044]

[0045] Among them, x′ is used to represent the normalized data value, x is used to represent the power trading data, μ is used to represent the mean of the power trading data, and σ is used to represent the standard deviation of the power trading data;

[0046] The weighted fusion processing subunit is used to perform weighted fusion on the normalized data by taking the following calculation formula to obtain the risk assessment data:

[0047]

[0048] Wherein, F is used to represent the risk assessment data, m is the number of data, and x i Represents the number of data to be fused, ω i ′ is used to represent data x i The corresponding normalized weight, ω i To represent the data x i The corresponding weights before normalization.

[0049] Preferably, the transaction network risk assessment unit includes: a risk situation ledger subunit and a risk reporting subunit;

[0050] The risk situation ledger subunit is used to evaluate the risk assessment data using a smart contract, calculate the risk score according to the three-level indicator system, and generate risk situation ledger data; and write the risk situation ledger data into a blockchain, and monitor the risk data and events on the blockchain through the security assessment framework;

[0051] The risk reporting subunit is used to generate a unique data hash value for each event based on the risk situation ledger data, and use the data hash value of each risk event as a leaf node, generate non-leaf nodes by combining the data hash values ​​of the leaf nodes, until the data hash value of the root node is calculated, thereby constructing a Merkle tree; and, store the root node data hash value of the Merkle tree on the blockchain as a verification set for the risk report, and generate risk report data including a summary of risk events, a root hash, and recommended measures.

[0052] Preferably, the recurrent neural network BiGRU model includes:

[0053] Reset gate: r t =σ(W r ·[h t-1 ,x t ]);

[0054] Update gate: z t =σ(Wz ·[h t-1 ,x t ]);

[0055] Hidden state: h~ t =tanh(W h ·[r t *h t-1 ,x t ]);

[0056] Output gate: h t =(1-z t )*h t-1 +z t *h~ t ;

[0057] Among them, r t is the reset gate, z t is the update gate, x t is the current input, h~ t is the current input and the previous hidden layer state summary, h t is the hidden layer output after updating the memory, W r is the weight matrix of the reset gate, W z is the weight matrix of the update gate, W h is the weight matrix of the hidden layer, σ is the Sigmoid function, and tanh is the hyperbolic tangent activation function.

[0058] Preferably, the intelligent risk response includes:

[0059] Automatically adjust access control rules, add identity verification steps, and limit access to high-risk areas;

[0060] When data integrity risks are detected, automatically adjust the frequency and scope of data backup and formulate a data recovery plan;

[0061] When the local policy is updated, the update event is automatically broadcast in the blockchain network;

[0062] When a node updates its defense strategy, the summary and parameters of the defense strategy are published in the blockchain network through smart contracts. Other nodes subscribe to the defense strategy update event and decide whether to adopt the update event based on their own security status and strategy compatibility.

[0063] It can be seen from the above technical scheme that in the power trading network risk assessment system based on blockchain and edge computing provided by the present invention, the security assessment framework construction module establishes a security assessment framework for risk assessment of power trading based on blockchain and edge computing, the indicator system construction module constructs a three-level indicator system for power trading network security risk assessment based on power trading data, the risk level assessment module conducts risk assessment on the power trading network based on the security assessment framework and the three-level indicator system, and applies the recurrent neural network BiGRU model to dynamically evaluate the risk level, and finally the risk response module performs intelligent risk response according to the dynamic evaluation result of the risk level. It can be seen that this scheme realizes the assessment of power trading network security based on the constructed three-level indicator system, and solves the problem of poor accuracy of risk assessment caused by the overly single evaluation indicators of the existing scheme. Moreover, the security assessment framework established by this scheme based on blockchain and edge computing is a distributed structure, and can dynamically evaluate the risks of the power trading network, so that this scheme can greatly improve the accuracy of power trading network security risk assessment and provide strong security protection for the power trading system. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Figure 1 A schematic diagram of a power trading network risk assessment system based on blockchain and edge computing provided in an embodiment of the present invention.

[0065] Figure 2 A schematic diagram of a security assessment framework building module provided by an embodiment of the present invention.

[0066] Figure 3 A schematic diagram of an indicator system construction module provided in an embodiment of the present invention.

[0067] Figure 4 A schematic diagram of a second-level evaluation index weight determination unit provided in an embodiment of the present invention.

[0068] Figure 5 A schematic diagram of a risk level assessment module provided in an embodiment of the present invention.

[0069] Figure 6 A schematic diagram of a transaction network risk assessment unit provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0070] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0071] Traditional network security risk assessment methods include qualitative, quantitative and a combination of qualitative and quantitative methods. In view of the characteristics of current power trading, there are mainly the following problems: (1) Existing network security risk assessment is centralized and static, which is not suitable for the distributed characteristics of power trading, and the static model cannot reflect system changes in real time; (2) The existing network security risk assessment indicator system cannot fully reflect the characteristics of power trading in the new power system. With the addition of new energy, the network security of power trading presents diversified characteristics, and the weights are different. The existing indicator system is too single; (3) The existing model has major deficiencies in data plasticity and non-tamperability; (4) With the expansion of the system, the computing overhead increases significantly.

[0072] As an emerging Internet computing, blockchain has the characteristics of decentralization, distributed block chain storage, security and reliability, and high transparency, providing a new solution in the field of power trading security risk assessment. At the same time, combined with edge computing, blockchain nodes and edge computing nodes are integrated to reduce bandwidth and storage costs, thereby establishing a dynamic assessment model for power trading network security risks in new power systems. Therefore, this solution considers combining the characteristics of power trading in new power systems, based on blockchain technology and edge computing, and combining a multi-dimensional indicator system to build a cloud-edge collaborative power trading network information security dynamic assessment model. Solve the problems existing in the current algorithm, such as centralized models, non-real-time evaluation, low data security, single indicator system, and high computational overhead. This solution can adapt to the development of new power systems, realize accurate, rapid, and real-time assessment of power trading network risks, and provide strong security guarantees for the entire system. For details, please see Figure 1-6 , the present invention provides a power trading network risk assessment system based on blockchain and edge computing, including: a security assessment framework construction module 10, an indicator system construction module 20, a risk level assessment module 30 and a risk response module 40;

[0073] A security assessment framework building module 10 is used to establish a security assessment framework for evaluating power transaction information based on blockchain and edge computing;

[0074] The indicator system construction module 20 is used to construct a three-level indicator system for power trading network security risk assessment based on power trading data; wherein the power trading data includes: market price data, transaction volume data, contract data, power generation and consumption data, settlement data and system status data;

[0075] A risk level assessment module 30 is used to conduct risk assessment on the power trading network based on the security assessment framework and the three-level indicator system, and to dynamically assess the risk level using a recurrent neural network BiGRU model;

[0076] The risk response module 40 is used to perform intelligent risk response according to the dynamic assessment results of the risk level.

[0077] In this embodiment, the security assessment of the power trading network is realized based on the constructed three-level indicator system, which solves the problem of poor risk assessment accuracy caused by the existing scheme's overly single evaluation indicators. In addition, the security assessment framework established by this scheme based on blockchain and edge computing is a distributed structure, and can dynamically assess the risks of the power trading network, so that this scheme can greatly improve the accuracy of the power trading network security risk assessment and provide strong security protection for the power trading system.

[0078] The security assessment framework building module 10 is used to establish a security assessment framework for evaluating power transaction information based on blockchain and edge computing; specifically, it may include a system requirement determination unit 11, a node deployment unit 12, a blockchain platform deployment unit 13, a smart contract deployment unit 14, a security configuration unit 15 and a monitoring mechanism configuration unit 16;

[0079] The system requirement determination unit 11 is used to determine the system's requirements for processing data; for example, the system's requirements for processing data may include requirements for processing speed and throughput of power trading data, requirements for data security, privacy and anti-tampering, the number of nodes that the distributed network needs to support, network topology and data synchronization requirements, etc.;

[0080] The node deployment unit 12 is used to determine the edge nodes that have sufficient computing power, storage capacity and network bandwidth and are compatible with the selected blockchain platform and related software; and to arrange the locations of the edge computing nodes according to the distribution of the power trading system;

[0081] The blockchain platform deployment unit 13 is used to determine a blockchain platform that meets the requirements of smart contract support, data security and scalability, such as Ethereum, Hyperledger Fabric, etc.; and install and configure blockchain nodes on each edge computing device, and set the communication protocol, port and network topology between nodes;

[0082] The smart contract deployment unit 14 is used to define the business logic of power transactions, such as transaction verification, settlement rules and automatic execution conditions, and compile and test the smart contract code; and deploy the smart contract to the blockchain to ensure that the smart contract is correctly executed on the blockchain node;

[0083] Security configuration unit 15, used for configuration of node authentication, data encryption and authority control, node testing and data synchronization testing;

[0084] The monitoring mechanism configuration unit 16 is used to establish a monitoring mechanism to track the operating status of the system in real time to achieve real-time monitoring and alarm monitoring.

[0085] In this embodiment, it is considered to deploy blockchain nodes on edge computing nodes to form a distributed structure, and each terminal device is given a unique blockchain identity; determine the needs of the power trading system, including data processing speed, security and distributed network requirements, and select edge computing devices suitable for deploying blockchain nodes; select a suitable blockchain platform and configure its operating environment on the edge node; deploy smart contracts on the blockchain to automatically execute power trading and related business logic; implement security configuration, including node authentication, data encryption, permission control, etc.; test the entire distributed structure to ensure the normal operation of each node, correct data synchronization, and system stability; establish a monitoring mechanism to track the system's operating status in real time, and thus propose a cloud-edge collaborative power trading information security assessment framework.

[0086] In one embodiment, the indicator system construction module 20 is used to construct a three-level indicator system for power trading network security risk assessment based on power trading data; wherein the power trading data includes: market price data, transaction volume data, contract data, power generation and consumption data, settlement data and system status data;

[0087] In this embodiment, market price data includes transaction price information of the electricity market; transaction volume data may include the amount of electricity traded in each time period; contract data may include the contract terms and conditions signed by the two parties to the transaction; power generation and consumption data may include power generation, power consumption and demand forecast data; settlement data may include payment and settlement records after the transaction is completed; system status data may include power grid operation status and equipment performance indicators.

[0088] Specifically, the indicator system construction module 20 may include: a third-level indicator determination unit 21, a second-level evaluation indicator weight determination unit 22, and a first-level evaluation indicator weight determination unit 23;

[0089] A three-level indicator determination unit 21, used to determine the evaluation indicators of each level constituting the three-level indicator system;

[0090] The second-level evaluation index weight determination unit 22 is used to construct a hesitant fuzzy preference relationship matrix according to the evaluation index, and to modify the hesitant fuzzy preference relationship matrix using the matrix consistency evaluation index; and to calculate the weight of each second-level evaluation index using the modified hesitant fuzzy preference relationship matrix;

[0091] The first-level evaluation index weight determination unit 23 is used to perform weighted summation on the weights of the second-level evaluation indexes of the same category to obtain the first-level evaluation index weights.

[0092] For the three-level indicator determination unit 21, the evaluation indicators at each level determined by it can be shown in the following table:

[0093]

[0094]

[0095] For the second-level evaluation index weight determination unit 22, it may include a preference relationship matrix construction subunit 221, a consistency check subunit 222, an expected relationship matrix conversion subunit 223, a limit matrix determination subunit 224 and a secondary index weight determination subunit 225;

[0096] The preference relationship matrix construction subunit 221 is used to construct the following hesitant fuzzy preference relationship matrix according to the index network hierarchy:

[0097]

[0098] Among them, h ij (p ij ) is used to represent the expert i's evaluation of the third-level evaluation index p i Relative to the evaluation index p j The preference degree of each element is i, j = 1, 2, ..., n, and the dimension of the H matrix is ​​n × n;

[0099] The consistency check subunit 222 is used to calculate the consistency value CR of the hesitant fuzzy preference relationship matrix by the following calculation formula; and, when CR is less than a preset value, modify the hesitant fuzzy preference relationship matrix until CR meets the requirement:

[0100]

[0101] Among them, λ is the eigenvalue of the hesitant fuzzy preference relationship matrix, RI is the random one-time index, CI is the statistical mean, and n is the matrix dimension;

[0102] In this embodiment, the CR value is used to describe the consistency of the hesitant fuzzy preference relationship matrix. When CR<0.1, it indicates that the matrix consistency is good, otherwise the matrix is ​​modified.

[0103] The expected relationship matrix conversion subunit 223 is used to convert the hesitant fuzzy preference relationship matrix corrected by the consistency check subunit 222 into an expected hesitant fuzzy preference relationship matrix by the following calculation formula:

[0104]

[0105] in, It is used to characterize the expected hesitant fuzzy preference relationship matrix, l is the loop variable, traversing each element in the modified hesitant fuzzy preference relationship matrix;

[0106] The limit matrix determination subunit 224 is used to calculate the expected hesitant fuzzy preference relationship matrix using the hierarchical analysis software to obtain a weighted supermatrix; and to perform limit calculation on the weighted supermatrix to obtain a limit matrix reflecting the dependency relationship between the elements of the network layer;

[0107] In this embodiment, the hierarchical analysis software may include professional software such as YAANP. The weighted supermatrix is ​​subjected to limit calculation, which can be calculated by the following formula:

[0108]

[0109] in, represents the limit matrix and W represents the weighted supermatrix.

[0110] The second-level indicator weight determination subunit 225 is used to obtain the weights of all second-level indicators according to the limit matrix.

[0111] The risk level assessment module 30 is used to conduct risk assessment on the power trading network based on the security assessment framework and the three-level indicator system, and dynamically assess the risk level using the recurrent neural network BiGRU model; it may specifically include: a data acquisition unit 31, a data processing unit 32, a trading network risk assessment unit 33 and a risk level assessment unit 34;

[0112] The data acquisition unit 31 is used to acquire power transaction data; wherein the power transaction data includes system data and external interface data;

[0113] In this embodiment, the collected power transaction data may include sensor data, system logs, access records, user activity records and other related data. In addition, it also includes data from different external interfaces involved in power transactions, such as power generation quotation systems, market information, SCADA systems, AGC systems, management information systems, financial / banking systems, other dispatching and trading centers, other market members, telephone systems, video systems, external business systems, etc. The data types involved are diverse.

[0114] The data processing unit 32 is used to clean the power transaction data to obtain risk assessment data;

[0115] Since the data types collected are diverse, it is necessary to clean the collected data to remove invalid and erroneous data and ensure the uniformity of the data format. The specific data processing unit 32 may include a normalization processing subunit and a weighted fusion processing subunit;

[0116] The normalization processing subunit is used to normalize the power transaction data through the following calculation formula:

[0117]

[0118] Among them, x′ is used to represent the normalized data value, x is used to represent the power trading data, μ is used to represent the mean of the power trading data, and σ is used to represent the standard deviation of the power trading data;

[0119] The weighted fusion processing subunit is used to perform weighted fusion on the normalized data by taking the following calculation formula to obtain risk assessment data:

[0120]

[0121] Among them, F is used to represent the risk assessment data, m is the number of data, and x i Represents the number of data to be fused, ω i ′ is used to represent data x i The corresponding normalized weight, ω i To represent the data x i The corresponding weights before normalization. The normalized weights can ensure that the sum of the weights is 1.

[0122] The transaction network risk assessment unit 33 is used to evaluate the risk assessment data using the smart contract configured on the security assessment framework, and calculate the risk score according to the three-level indicator system to obtain the risk situation ledger data; and generate risk report data from the risk situation ledger data of the power transaction based on the Merkle tree;

[0123] In this embodiment, it is considered to generate a computer communication network security risk situation ledger based on the smart contract technology of the blockchain, and on this basis, a risk report is generated based on the Merkle tree; the Merkle tree can accurately track the dynamic changes of the risk situation and analyze the changes in risks in real time; in the process of generating the computer communication network risk report, the power transaction network security risk association tree is established based on the blockchain Merkle tree, the corresponding association between risks is completed, and the risk information location retrieval is realized efficiently, reliably, and arbitrarily. Specifically, the trading network risk assessment unit 33 includes: a risk situation ledger subunit 331 and a risk report subunit 332;

[0124] The risk situation ledger subunit 331 is used to evaluate the risk assessment data using smart contracts, calculate the risk score according to the three-level indicator system, and generate risk situation ledger data; and write the risk situation ledger data into the blockchain, and monitor the risk data and events on the blockchain through the security assessment framework;

[0125] The risk reporting subunit 332 is used to generate a unique data hash value for each event based on the risk situation ledger data, and use the data hash value of each risk event as a leaf node, generate non-leaf nodes by combining the data hash values ​​of the leaf nodes, until the data hash value of the root node is calculated, thereby constructing a Merkle tree; and, store the root node data hash value of the Merkle tree on the blockchain as a verification set for the risk report, and generate risk report data containing a summary of risk events, a root hash, and recommended measures.

[0126] In this embodiment, it specifically includes (1) generating a power trading network security risk situation ledger based on the smart contract technology of the blockchain to achieve data transparency and immutability. First, submit the data to the blockchain through a smart contract; use the smart contract to automatically evaluate the data and calculate the risk score based on the three-level indicator system; generate a risk situation ledger report and write it to the blockchain. Here, the ledger data is queried through a blockchain browser or a custom application to ensure the integrity and authenticity of the ledger data; establish a real-time monitoring system to monitor the risk data and events on the blockchain; and regularly audit the ledger data to ensure the accuracy and compliance of the risk data records. (2) Based on the Merkle tree, generate a risk report from the power trading risk situation ledger. The Merkle tree is a hash tree structure that can efficiently verify data. This method combines the transparency of the blockchain and the data integrity guarantee of the Merkle tree to generate and verify the risk report. The power trading risk situation ledger data is entered into the system, and a unique hash value is generated for each event; the data hash value of each risk event is used as a leaf node, and non-leaf nodes are generated by combining the hash values ​​of the leaf nodes until the hash value of the root node is calculated, thereby constructing a Merkle tree. The root hash of the Merkle tree is stored on the blockchain as a fingerprint of the risk report. The root hash of the Merkle tree is calculated locally or on the blockchain and compared with the stored root hash to verify the integrity of the data; a report containing a summary of risk events, root hash, recommended actions, etc. is generated

[0127] The risk level assessment unit 34 is used to dynamically assess the risk level using the historical data and the current risk report data and the recurrent neural network BiGRU model. In this process, the recurrent neural network BiGRU model corresponding to time t is:

[0128] Reset gate: r t =σ(W r ·[h t-1 ,x t ]);

[0129] Update gate: z t =σ(W z ·[h t-1 ,x t ]);

[0130] Hidden state: h~ t =tanh(W h ·[r t *h t-1 ,x t ]);

[0131] Output gate: h t =(1-z t )*h t-1 +z t *h~ t ;

[0132] Among them, r t is the reset gate, z t is the update gate, x t is the current input, h~ t is the current input and the previous hidden layer state summary, h t is the hidden layer output after updating the memory, W r is the weight matrix of the reset gate, W z is the weight matrix of the update gate, W h is the weight matrix of the hidden layer, σ is the Sigmoid function, and tanh is the hyperbolic tangent activation function.

[0133] In this embodiment, BiGRU performs distributed training to accelerate the model training process and improve the ability to process large-scale data. Data parallelism is implemented here. The entire model is replicated on multiple edge computing nodes, and each node processes a different subset of data. After each iterative training, each node calculates the gradient, and then calculates the merged gradient through the parameter aggregation method in federated learning, and updates the model parameters. PyTorch provides the torch.distributed package, which supports multi-GPU and multi-node training. The smart contract in the application blockchain triggers the model update, that is, according to the set smart contract rules, the model update process is automatically started based on the trigger conditions.

[0134] The risk response module 40 is used to perform intelligent risk response according to the dynamic assessment results of the risk level. Specifically, it can include automatically adjusting access control rules, adding identity authentication steps, limiting access rights to high-risk areas, etc.; when data integrity risks or vulnerabilities are detected, automatically adjusting the frequency and scope of data backup, and formulating a data recovery plan; other nodes in the blockchain share response strategies, including defining strategy sharing and update mechanisms through smart contracts, and automatically broadcasting strategy update events in the blockchain network when local strategies are updated. When a node updates its defense strategy, the summary and parameters of the defense strategy are published in the blockchain network through smart contracts, and other nodes subscribe to defense strategy update events and decide whether to adopt the update events based on their own security status and strategy compatibility.

[0135] The present specification also provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed in a computer, the computer is caused to execute a method in any one of the embodiments in the specification.

[0136] The present specification also provides a computing device, including a memory and a processor, wherein the memory stores executable code, and when the processor executes the executable code, the method in any embodiment of the specification is implemented.

[0137] The modules or units in the device of the embodiment of the present invention can be combined, divided and deleted according to actual needs. The above disclosure is only the preferred embodiment of the present invention, and of course it cannot be used to limit the scope of the rights of the present invention. Those skilled in the art can understand that all or part of the processes of the above embodiment are implemented, and the equivalent changes made according to the claims of the present invention still fall within the scope of the invention.

Claims

1. A power trading network risk assessment system based on blockchain and edge computing, characterized in that: include: Safety assessment framework building module, indicator system building module, risk level assessment module and risk response module; The security assessment framework building module is used to establish a security assessment framework for evaluating power transaction information based on blockchain and edge computing; The indicator system construction module is used to construct a three-level indicator system for power trading network security risk assessment based on power trading data; wherein the power trading data includes: market price data, transaction volume data, contract data, power generation and consumption data, settlement data and system status data; The risk level assessment module is used to perform risk assessment on the power trading network based on the security assessment framework and the three-level indicator system, and dynamically assess the risk level using a recurrent neural network BiGRU model; The risk response module is used to perform intelligent risk response according to the dynamic assessment results of the risk level.

2. The power trading network risk assessment system based on blockchain and edge computing according to claim 1 is characterized in that: The security assessment framework building module includes: a system requirement determination unit, a node deployment unit, a blockchain platform deployment unit, a smart contract deployment unit, a security configuration unit and a monitoring mechanism configuration unit; The system requirement determination unit is used to determine the system's requirements for processing data; wherein the system's requirements for processing data include: requirements for processing speed and throughput of power trading data, requirements for data security, privacy and anti-tampering, and the number of nodes that the distributed network needs to support, network topology and data synchronization requirements; The node deployment unit is used to determine edge nodes that have sufficient computing power, storage capacity, and network bandwidth and are compatible with the selected blockchain platform and related software; and to arrange the locations of edge computing nodes according to the distribution of the power trading system; The blockchain platform deployment unit is used to determine a blockchain platform that meets the requirements of smart contract support, data security and scalability; and install and configure blockchain nodes on each edge computing device, and set communication protocols, ports and network topology between nodes; The smart contract deployment unit is used to define the business logic of the power transaction and compile it into a smart contract code; and deploy the smart contract to the blockchain to ensure that the smart contract is correctly executed on the blockchain node; The security configuration unit is used to configure node authentication, data encryption and authority control; The monitoring mechanism configuration unit is used to establish a monitoring mechanism to track the operating status of the system in real time.

3. The power trading network risk assessment system based on blockchain and edge computing according to claim 1 is characterized in that: The indicator system construction module includes: a third-level indicator determination unit, a second-level evaluation indicator weight determination unit and a first-level evaluation indicator weight determination unit; The three-level indicator determination unit is used to determine the evaluation indicators of each level constituting the three-level indicator system; The second-level evaluation index weight determination unit is used to construct a hesitant fuzzy preference relationship matrix according to the evaluation index, and to modify the hesitant fuzzy preference relationship matrix using the matrix consistency evaluation index; and to calculate the weight of each second-level evaluation index using the modified hesitant fuzzy preference relationship matrix; The first-level evaluation index weight determination unit is used to perform weighted summation on the weights of the second-level evaluation indexes of the same category to obtain the first-level evaluation index weights.

4. The power trading network risk assessment system based on blockchain and edge computing according to claim 3 is characterized in that: The evaluation indicators at each level are shown in the following table:

5. The power trading network risk assessment system based on blockchain and edge computing according to claim 3 is characterized in that: The second-level evaluation index weight determination unit includes: a preference relationship matrix construction subunit, a consistency verification subunit, an expected relationship matrix conversion subunit, a limit matrix determination subunit and a secondary index weight determination subunit; The preference relationship matrix construction subunit is used to construct the following hesitant fuzzy preference relationship matrix according to the indicator network hierarchy: Among them, h ij (p ij ) is used to represent the expert i's evaluation of the third-level evaluation index p i Relative to the evaluation index p j The preference degree of each element is i, j = 1, 2, ..., n, and the dimension of the H matrix is ​​n × n; The consistency check subunit is used to calculate the consistency value CR of the hesitant fuzzy preference relationship matrix by the following calculation formula; and, when CR is less than a preset value, modify the hesitant fuzzy preference relationship matrix until CR meets the requirement: Among them, λ is the eigenvalue of the hesitant fuzzy preference relationship matrix, RI is the random one-time index, CI is the statistical average, and n is the dimension of the H matrix; The expected relationship matrix conversion subunit is used to convert the hesitant fuzzy preference relationship matrix corrected by the consistency check subunit into the expected hesitant fuzzy preference relationship matrix through the following calculation formula: in, It is used to characterize the expected hesitant fuzzy preference relationship matrix, l is a loop variable, and iterates over each element in the modified hesitant fuzzy preference relationship matrix; The limit matrix determination subunit is used to calculate the expected hesitant fuzzy preference relationship matrix using hierarchical analysis software to obtain a weighted supermatrix; and to perform limit calculation on the weighted supermatrix to obtain a limit matrix reflecting the dependency relationship between the elements of the network layer; The secondary indicator weight determination subunit is used to obtain the weights of all second-level indicators according to the limit matrix.

6. The power trading network risk assessment system based on blockchain and edge computing according to claim 1 is characterized in that: The risk level assessment module includes: a data acquisition unit, a data processing unit, a transaction network risk assessment unit and a risk level assessment unit; The data acquisition unit is used to acquire power transaction data; wherein the power transaction data includes system data and external interface data; The data processing unit is used to clean the power transaction data to obtain risk assessment data; The transaction network risk assessment unit is used to evaluate the risk assessment data using the smart contract configured on the security assessment framework, and calculate the risk score according to the three-level indicator system to obtain risk situation ledger data; and generate risk report data from the risk situation ledger data of power transactions based on the Merkle tree; The risk level assessment unit is used to dynamically assess the risk level by applying the recurrent neural network BiGRU model through historical data and current risk report data.

7. The power trading network risk assessment system based on blockchain and edge computing according to claim 6 is characterized in that: The data processing unit includes a normalization processing subunit and a weighted fusion processing subunit; The normalization processing subunit is used to perform normalization processing on the power transaction data by the following calculation formula: Among them, x′ is used to represent the normalized data value, x is used to represent the power trading data, μ is used to represent the mean of the power trading data, and σ is used to represent the standard deviation of the power trading data; The weighted fusion processing subunit is used to perform weighted fusion on the normalized data by taking the following calculation formula to obtain the risk assessment data: Wherein, F is used to represent the risk assessment data, m is the number of data, and x i Represents the number of data to be fused, ω i ′ is used to represent data x i The corresponding normalized weight, ω i To represent the data x i The corresponding weights before normalization.

8. The power trading network risk assessment system based on blockchain and edge computing according to claim 6 is characterized in that: The transaction network risk assessment unit includes: a risk situation ledger subunit and a risk reporting subunit; The risk situation ledger subunit is used to evaluate the risk assessment data using a smart contract, calculate the risk score according to the three-level indicator system, and generate risk situation ledger data; and write the risk situation ledger data into a blockchain, and monitor the risk data and events on the blockchain through the security assessment framework; The risk reporting subunit is used to generate a unique data hash value for each event based on the risk situation ledger data, and use the data hash value of each risk event as a leaf node, generate non-leaf nodes by combining the data hash values ​​of the leaf nodes, until the data hash value of the root node is calculated, thereby constructing a Merkle tree; and, store the root node data hash value of the Merkle tree on the blockchain as a verification set for the risk report, and generate risk report data including a summary of risk events, a root hash, and recommended measures.

9. The power trading network risk assessment system based on blockchain and edge computing according to claim 6 is characterized in that: The recurrent neural network BiGRU model includes: Reset gate: r t =σ(W r ·[h t-1 ,x t ]); Update gate: z t =σ(W z ·[h t-1 ,x t ]); Hidden state: h~ t =tanh(W h ·[r t *h t-1 ,x t ]); Output gate: h t =(1-z t )*h t-1 +z t *h~ t ; Among them, r t is the reset gate, z t is the update gate, x t is the current input, h~ t is the current input and the previous hidden layer state summary, h t is the hidden layer output after updating the memory, W r is the weight matrix of the reset gate, W z is the weight matrix of the update gate, W h is the weight matrix of the hidden layer, σ is the Sigmoid function, and tanh is the hyperbolic tangent activation function.

10. The power trading network risk assessment system based on blockchain and edge computing according to claim 1 is characterized in that: The intelligent risk response includes: Automatically adjust access control rules, add identity verification steps, and limit access to high-risk areas; When data integrity risks are detected, data backup frequency and scope are automatically adjusted and data recovery plans are formulated; When the local policy is updated, the update event is automatically broadcast in the blockchain network; When a node updates its defense strategy, the summary and parameters of the defense strategy are published in the blockchain network through a smart contract. Other nodes subscribe to the defense strategy update event and decide whether to adopt the update event based on their own security status and strategy compatibility.

Images