Cross-chain storage system for electronic accounting documents in supply chain finance

Abstract

The invention discloses a cross-chain storage system for electronic accounting documents in supply chain finance, and relates to the technical field of supply chain finance and block chains. The system comprises a certificate data acquisition module, a multi-dimensional feature extraction module, a block chain alliance module, a cross-chain communication protocol module and the like. The multi-dimensional feature extraction module is used for mining and fusing features such as time and amount of the electronic accounting document; the block chain alliance module is utilized to construct a heterogeneous block chain system, and the cross-chain communication protocol module is combined to realize cross-chain data transmission; the consensus efficiency and security are improved by means of a distributed consensus verification module; efficient management of certificate storage is achieved through the certificate storage management module, and a data security enhancement module, a privacy protection module and the like are further arranged. According to the invention, the problems of lack of cross-chain collaboration and the like in the existing scheme are solved, efficient and safe cross-chain evidence storage adaptive to complex scenes is realized, and the supply chain financial business collaboration efficiency and the data security are improved.

Description

Technical Field

[0001] The present invention relates to the fields of supply chain finance and blockchain technology, and specifically to a cross-chain evidence storage system for electronic accounting vouchers in supply chain finance. Background Art

[0002] In the supply chain finance business process, the storage of electronic accounting vouchers is a key step in ensuring transaction credibility and achieving risk management. The traditional storage model relies on a centralized database, where management personnel manually enter and maintain voucher information. Financial and risk control personnel are required to verify the authenticity, relevance, and compliance of each voucher with business rules. This manual storage method is not only inefficient, with the collection and on-chain storage of a single voucher taking up to several hours, but is also prone to data tampering and loss due to human error or external attacks. According to industry statistics, the error or loss rate of voucher data in the centralized storage model is approximately 2%-4%. In complex supply chain scenarios involving multiple levels of suppliers and cross-platform collaboration (such as cross-border procurement and multi-level distribution), data consistency and security issues are even more prominent.

[0003] Currently, electronic accounting voucher storage solutions using blockchain technology are gradually emerging. The mainstream system generally consists of a data collection module, a single-chain storage module, and a basic verification component. The typical workflow is as follows:

[0004] 1. Voucher collection: Capture electronic accounting voucher files from the business system through the interface and extract key fields (such as transaction amount, participating entities, and business time);

[0005] 2. Single-chain evidence storage: Store the credential hash value or simplified data in a single blockchain, and use the blockchain's tamper-proof nature to retain records;

[0006] 3. Manual verification supplement: For certificates involving cross-chain interactions and complex business relationships, manual intervention is still required to confirm the compatibility of the evidence storage logic with the business.

[0007] However, the existing blockchain evidence storage system has obvious limitations:

[0008] 1. Lack of cross-chain collaboration: Only a single blockchain is supported for evidence storage. Heterogeneous blockchains (such as consortium chains and private chains) used by different supply chain participants cannot effectively interact with each other, making it difficult to achieve cross-platform and cross-institutional evidence storage and management throughout the entire life cycle of credentials, resulting in data silos.

[0009] 2. Insufficient feature utilization: The multi-dimensional features of electronic accounting documents (such as time-related features and multi-party relationship features) are not fully explored, and the value of the stored data is not effectively released, making it difficult to support in-depth risk control and business analysis based on the document data.

[0010] 3. Consensus and security shortcomings: The consensus mechanism used has poor fault tolerance and efficiency adaptability in cross-chain scenarios. Faced with the dynamic changes in nodes and high-frequency data interactions in complex supply chain networks, it is prone to consensus delays and verification vulnerabilities, and is unable to accurately identify risks such as credential tampering and forgery.

[0011] 4. Simple management functions: There is a lack of efficient tools and standardized processes for data retrieval, updating, and auditing after evidence storage, making it difficult to quickly locate historical vouchers and process changes in voucher information. It is also impossible to provide a complete and clear evidence tracking track for audits.

[0012] Therefore, in view of this, the existing technology was studied and improved, and a cross-chain notarization system for electronic accounting vouchers in supply chain finance was proposed. Summary of the Invention

[0013] The technical problem to be solved by the present invention is to solve the problems of lack of cross-chain collaboration, insufficient feature utilization, shortcomings in consensus and security, and simple evidence management functions in the existing supply chain finance electronic accounting voucher evidence storage scheme, so as to achieve efficient, secure and adaptable cross-chain evidence storage for complex scenarios.

[0014] The technical solution adopted by the present invention is: a cross-chain evidence storage system for electronic accounting documents in supply chain finance, comprising:

[0015] Voucher data collection module, used to collect electronic accounting voucher data from the supply chain finance system;

[0016] Multi-dimensional feature extraction module, used to extract the multi-dimensional feature vector of the electronic accounting document

[0017] Blockchain alliance module, including multiple heterogeneous blockchain systems B={B1,B2,…,B m};

[0018] Cross-chain communication protocol module, which realizes cross-chain data transmission based on improved hash time lock contract (HTLC) and side chain technology;

[0019] Distributed consensus verification module, using Byzantine fault-tolerant consensus algorithm in is the transaction set, is the set of verification nodes, α is the fault tolerance threshold;

[0020] The voucher storage management module is used to retrieve, verify and update the stored electronic accounting vouchers.

[0021] As a further solution of the present invention: the multi-dimensional feature extraction module further includes:

[0022] Time dimension feature extraction unit, extracting the voucher generation time tg , business occurrence time t b and the evidence storage time t s , and construct the time feature vector

[0023] Amount dimension feature extraction unit, extracts voucher amount A, currency C and exchange rate E, and constructs amount feature vector

[0024] Transaction relationship dimension feature extraction unit, extracting the identity information P1, P2 of both parties and the transaction type T t , and construct the transaction relationship feature vector

[0025] Feature fusion unit, through the formula The multi-dimensional feature vectors are fused, where σ is the activation function, W is the weight matrix, and b is the bias vector.

[0026] As a further solution of the present invention: In the blockchain alliance module, each blockchain system B i With different block structures S i 、Consensus Mechanism M i and smart contract engine E i , and meet the heterogeneous compatibility conditions:

[0027]

[0028] where φ ij is the block structure conversion function, and satisfies:

[0029] H(φ ij (x))=H(x)mod p

[0030] Here H is a hash function and p is a secure prime number.

[0031] As a further solution of the present invention: the cross-chain communication protocol module adopts a cross-chain data verification mechanism based on zero-knowledge proof, specifically including:

[0032] Data commitment generation unit, generates Merkle tree commitment C of electronic accounting document data D D =MerkleTree(D);

[0033] Zero-knowledge proof generation unit generates proof π, which enables the verifier to verify C without revealing the specific content of D D effectiveness;

[0034] The cross-chain data transfer unit is calculated by the formula:

[0035] Transfer(CD ,π,B i ,B j )=Verify(π,C D )∧Relay(C D ,B j )

[0036] Realize from blockchain B i To blockchain B j , where Verify is the verification function and Relay is the relay function.

[0037] As a further solution of the present invention: the Byzantine fault-tolerant consensus algorithm adopted by the distributed consensus verification module The following steps are involved:

[0038] Pre-preparation phase: the master node broadcasts the pre-preparation message Where v is the view number, n is the sequence number, and d is Abstract;

[0039] Preparation phase: Verification node Send Prepare Message<PREPARE,v,n,d,i> , when receiving at least 2f+1 different verification nodes’ prepare messages, the verification node enters the commit phase, where

[0040] Commit phase: Verification node sends commit message<COMMIT,v,n,d,i> , when receiving commit messages from at least 2f+1 different verification nodes, the verification node executes the transaction and updates the status;

[0041] The algorithm meets the security conditions:

[0042]

[0043] and the active condition:

[0044]

[0045] As a further solution of the present invention: the credential storage management module further includes:

[0046] The evidence index construction unit builds an evidence index structure based on Bloom filter I = BloomFilter({h(D1),h(D2),…,h(D k )}), where D i is the i-th electronic accounting document, h is the hash function;

[0047] Evidence retrieval unit, through the formula:

[0048]

[0049] Achieve fast retrieval of stored evidence, where q is the retrieval query;

[0050] Evidence verification unit, through the formula:

[0051] Verify(D,proof)=H(D)=extractRoot(proof)

[0052] Verify the integrity of the proof, where proof is the proof of the proof and extractRoot is the function that extracts the root hash from the proof;

[0053] The evidence update unit uses the version control mechanism to generate a new evidence for the updated certificate D′<D′,v+1,timestamp,signature> , and establish an association with the original evidence.

[0054] As a further solution of the present invention: also include:

[0055] The data security enhancement module uses the homomorphic encryption algorithm E and the secure multi-party computing protocol to achieve:

[0056] E(f(x1,x2,…,x n ))=f(E(x1),E(x2),…,E(x n ))

[0057] Where f is any polynomial time computable function;

[0058] The privacy protection module uses the zero-knowledge proof protocol Π = (Prover, Verifier) ​​to enable the verifier to verify the validity of the statement stmt without leaking any additional information:

[0059]

[0060] where ≈ c Denotes computational indistinguishability.

[0061] As a further solution of the present invention: the electronic accounting document has a multi-dimensional time stamp structure TS = [t1, t2, ..., t k ],in:

[0062] t1 is the timestamp of the transaction, satisfying t1 = Hash(D||nonce1);

[0063] t2 is the timestamp for generating the credential, satisfying t2 = Hash(t1||D′||nonce2);

[0064] t3 is the timestamp of the first proof storage, satisfying t3 = Hash(t2||B1||nonce3);

[0065] t i The timestamp of the i-th cross-chain evidence storage, satisfying t i =Hash(t i-1 ||B i-1 ||B i ||nonce i );

[0066] The timestamp sequence satisfies the chain verification relationship:

[0067]

[0068] The context i Context information generated for the i-th timestamp.

[0069] As a further solution of the present invention, the system further includes a risk assessment module, which performs risk assessment on the electronic accounting voucher through the following steps:

[0070] Construct the credential feature matrix X = [x ij ] n×m , where x ij represents the jth feature of the i-th voucher;

[0071] Train the risk assessment model M so that:

[0072] M(X)=σ(W L ·σ(W L-1 ·…·σ(W1·X+b1)…+b L-1 )+b L )

[0073] Where W l and b l are the weight matrix and bias vector of the lth layer respectively, and σ is the activation function;

[0074] Calculating risk scores in is the feature weight vector;

[0075] According to the risk score R and the preset threshold θ, the credentials are divided into high risk (R>θ), medium risk (R∈[θ / 2,θ]) and low risk (R<θ / 2) categories.

[0076] As a further solution of the present invention, the system further includes an audit tracking module, which implements audit tracking of the entire life cycle of electronic accounting documents through the following steps:

[0077] Construct a credential event graph G = (V, E), where the node set V = {v1, v2, ..., v n} represents the credential state, and the edge set E={e1,e2,…,e m} indicates a state transition event;

[0078] For each event e i =(v j ,v k ,t i ,a i )Record timestamp t i and operating subject a i ;

[0079] Define the audit path as a node sequence P = [v1, v2, ..., v k ],satisfy

[0080] Constructing Audit Proof π P , such that:

[0081]

[0082] The proof is based on zero-knowledge proof technology to ensure the privacy and security of the audit process.

[0083] Beneficial effects of the present invention:

[0084] 1. Efficient cross-chain collaboration: Break through the limitations of single blockchain evidence storage, build a heterogeneous blockchain alliance and cross-chain communication protocol, achieve smooth interaction between the blockchain systems of different supply chain participants, break down data silos, and make electronic accounting vouchers more coherent in the entire life cycle evidence management in cross-platform and cross-institutional scenarios, thereby improving the collaborative efficiency of supply chain financial business.

[0085] 2. Deeply unlock the value of features: The multi-dimensional feature extraction and fusion module comprehensively mines the time, amount, transaction relationship and other features of electronic accounting vouchers, providing richer data support for supply chain financial risk control and business analysis, helping to accurately identify business models and potential risks, and optimize financial decision-making.

[0086] 3. Consensus security adaptation and upgrade: The improved distributed consensus verification algorithm is adapted to cross-chain scenarios, enhancing fault tolerance and efficiency. In the face of dynamic changes in supply chain network nodes and high-frequency data interactions, it can effectively avoid consensus delays, accurately identify the risks of credential tampering and forgery, and ensure the security and reliability of stored evidence data.

[0087] 4. Intelligent and Improved Evidence Management: Optimized evidence indexing, retrieval, verification, and update functions, combined with risk assessment and audit tracking modules, make data management after evidence storage more efficient and intelligent. This allows for rapid location of historical vouchers and standardized processing of information changes, providing a clear and complete evidence track for audits and improving the compliance and traceability of supply chain finance services. BRIEF DESCRIPTION OF THE DRAWINGS

[0088] Figure 1 This is a framework diagram of a cross-chain evidence storage system for electronic accounting documents in supply chain finance according to the present invention.

[0089] Figure 2 This is a flow chart of the multi-dimensional feature extraction module of a cross-chain evidence storage system for electronic accounting vouchers in supply chain finance according to the present invention.

[0090] Figure 3 This is a flow chart of the cross-chain communication protocol module of a cross-chain evidence storage system for electronic accounting vouchers in supply chain finance according to the present invention.

[0091] Figure 4 This is a flow chart of the distributed consensus verification module of a cross-chain evidence storage system for electronic accounting vouchers in supply chain finance according to the present invention.

[0092] Figure 5 This is a flow chart of the voucher evidence management module of a cross-chain evidence storage system for electronic accounting vouchers in supply chain finance according to the present invention.

[0093] Figure 6 This is a flow chart of the risk assessment module of a cross-chain evidence storage system for electronic accounting vouchers in supply chain finance according to the present invention.

[0094] Figure 7 This is a flow chart of the audit tracking module of a cross-chain evidence storage system for electronic accounting documents in supply chain finance according to the present invention. DETAILED DESCRIPTION

[0095] The present invention will be further described below.

[0096] See also Figure 1-7

[0097] Example 1: Cross-chain storage application of electronic accounting vouchers based on the automotive supply chain

[0098] Application Scenario

[0099] An automotive manufacturing supply chain encompasses core enterprises, multi-tier suppliers, and financial institutions, with each participant utilizing heterogeneous blockchains (e.g., consortium and private blockchains). Under traditional evidence storage models, storing a single document takes approximately four hours, with a data error rate of up to 3%, resulting in inefficient cross-institutional collaboration.

[0100] Specific implementation steps

[0101] 1. Multi-dimensional feature extraction and fusion

[0102] Time dimension: Extract purchase order signing time t g , invoice generation time t b , first evidence storage time t s , construct the time feature vector:

[0103]

[0104] Amount dimension: Extract invoice amount A = 100,000 yuan, currency C = CNY, exchange rate E = 1, and construct the amount feature vector:

[0105]

[0106] Transaction relationship dimension: Extract the identity information of the core enterprise P1 and the first-tier supplier P2, and the transaction type T t =Raw material procurement, constructing transaction relationship feature vector:

[0107]

[0108] Feature fusion: Features are fused through the activation function σ (such as ReLU), the weight matrix W and the bias vector b:

[0109]

[0110] 2. Cross-chain communication and consensus verification

[0111] Cross-chain data transfer: Based on the improved Hash Time Lock Contract (HTLC), the supplier stores the invoice hash value on the private chain B1 and then transfers it to the core enterprise alliance chain B2 through the following formula:

[0112] Transfer(C D ,π,B1,B2)=Verify(π,C D )∧Relay(C D ,B2)

[0113] Among them C D =MerkleTree(D) is the Merkle tree commitment of the credential data D, and π is the zero-knowledge proof.

[0114] Byzantine Fault Tolerant Consensus: Validation Node Set Contains 101 nodes, fault tolerance threshold After the master node broadcasts the prepare message, the verification node needs to collect at least 2f+1=67 prepare messages before entering the commit phase.

[0115] 3. Evidence management and risk assessment

[0116] Evidence index and retrieval: Build index I based on Bloom filter, and retrieve the certificate data D i Calculate the hash value h(D i ):

[0117] I=BloomFilter(h(D1),h(D2),…,h(D k ))

[0118] When retrieving query q, if h(q)∈I, then return “may exist”.

[0119] Risk Assessment Model: Constructing the Feature Matrix Calculate the risk score R through a multi-layer neural network:

[0120] M(X)=σ(W L ·σ(W L-1 ·····σ(W1·X+b1)…+b L-1 )+b L )

[0121]

[0122] in is the feature weight vector, the preset risk threshold θ=0.7, and if R>0.7, it is judged as high risk.

[0123] Implementation Effect

[0124] 1. Cross-chain evidence storage time is shortened to 30 minutes, and the data error rate is reduced to 0.5%;

[0125] 2. Based on time feature vector Analysis revealed that the risk probability of a voucher where the time between supplier invoice generation and transaction occurrence exceeds three days increases by 20%.

[0126] 3. Through Byzantine fault-tolerant consensus, the consistency of evidence can still be guaranteed even if 30% of the nodes are maliciously tampered with.

[0127] Example 2: Multi-currency cross-chain evidence storage for cross-border e-commerce supply chains

[0128] Application Scenario

[0129] A cross-border e-commerce platform connects domestic suppliers, overseas distributors, and cross-border payment institutions, involving transactions in RMB (CNY) and Euro (EUR). Under the traditional evidence storage model, the multi-currency reconciliation error rate reaches 5%, and cross-chain data verification takes more than 2 hours.

[0130] Specific implementation steps

[0131] 1. Multi-dimensional timestamp construction

[0132] Business occurrence timestamp (order placement):

[0133] t1=Hash(D||nonce1)

[0134] Voucher generation timestamp (invoice issuance):

[0135] t2=Hash(t1||D′||nonce2)

[0136] Cross-chain evidence timestamp (from domestic alliance chain B1 to overseas public chain B2):

[0137] t i =Hash(t i-1 ||B1||B2||nonce i )

[0138] 2. Heterogeneous blockchain compatibility mechanism

[0139] Block structure conversion function φ between domestic alliance chain B1 and overseas public chain B2 12 satisfy:

[0140] H(φ 12 (x))=H(x)mod p

[0141] Where H is the SHA-256 hash function, and the secure prime number p = 2 256 -2 32 -977.

[0142] 3. Evidence verification and update

[0143] Integrity verification: Compare the original certificate hash value with the stored root hash value:

[0144] Verify(D,proof)=H(D)=extractRoot(proof)

[0145] Version update: When the overseas distributor invoice amount is changed to D′=15000EUR, a new certificate is generated:

[0146] <D′,v+1,timestamp,signature>

[0147] The version number v is updated from 1 to 2, and the timestamp timestamp contains the cross-chain operation time.

[0148] Implementation Effect

[0149] 1. The error rate of multi-currency evidence reconciliation has been reduced to 0.3%, and the cross-chain verification time has been shortened to 10 minutes;

[0150] 2. Based on the amount feature vector Analysis shows that when the EUR / CNY exchange rate fluctuates by more than 5%, the risk of certificate tampering increases by 15%;

[0151] 3. Through the zero-knowledge proof protocol π, the account privacy of both parties to the transaction is protected when verifying cross-border payment credentials.

[0152] Example 3: Warehouse Receipt Pledge Cross-Chain Evidence Storage in the Agricultural Products Supply Chain

[0153] Application Scenario

[0154] A certain agricultural product supply chain includes farmers, cooperatives, and processing companies. There is a risk of document forgery in warehouse receipt pledge financing. Under the traditional model, warehouse receipt verification requires manual verification for 3 hours, and the forgery rate reaches 2.8%.

[0155] Specific implementation steps

[0156] 1. Distributed consensus and security enhancement

[0157] Byzantine Fault Tolerant Consensus Algorithm Steps:

[0158] The master node broadcasts the pre-prepare message:

[0159]

[0160] After receiving 2f+1=41 prepare messages, the verification node enters the submission phase (the total number of nodes f = 40). Homomorphic encryption protection: Encrypted calculation of warehouse receipt quantity x = 1000 tons and unit price y = 5000 yuan / ton:

[0161] E(x·y)=E(x)·E(y)

[0162] Financial institutions can verify the value of warehouse receipts in encrypted form without decrypting the original data.

[0163] 2. Audit Trail and Event Graph Construction

[0164] In the voucher event graph G = (V, E), node v1 represents "warehouse receipt generation", node v2 represents "pledge registration", and edge e1 = (v1, v2, t = 2025-06-01, a = cooperative) records the state transition timestamp t and the operation subject a.

[0165] Audit Path Verification:

[0166]

[0167] Implementation Effect

[0168] 1. Warehouse receipt verification time is shortened to 15 minutes, and the counterfeiting rate is reduced to 0.1%;

[0169] 2. Through the audit tracking module, it was discovered that a cooperative had transferred goods without authorization after pledging warehouse receipts. The risk score was R = 0.85 > θ = 0.7, and a timely warning was issued.

[0170] 3. Homomorphic encryption computing efficiency is increased by 30%, meeting the high-frequency trading needs of the agricultural product supply chain.

[0171] Example 4: Cross-chain evidence storage of construction industry supply chain progress payments

[0172] Application Scenario

[0173] A construction supply chain includes owners, general contractors, and subcontractors. Progress payment vouchers are subject to delayed confirmation. Under the traditional model, the voucher lag rate reaches 18%, impacting financing efficiency.

[0174] Specific implementation steps

[0175] 1. Risk Assessment and Characteristic Matrix

[0176] Construct the feature matrix X, where x i1 is the amount of the i-th voucher, x i2 The project completion rate is:

[0177]

[0178] Risk score calculation: If the project completion rate is <0.5 and the amount is >500,000 yuan, the weight vector but:

[0179]

[0180] Determined to be high risk (R>0.7).

[0181] 2. Cross-chain data commitment generation

[0182] Merkle tree commitment of progress payment voucher D:

[0183] C D =MerkleTree(D1,D2,…,D n )

[0184] Among them D i The root hash is used for cross-chain verification.

[0185] Implementation Effect

[0186] 1. The document lag rate has been reduced to 3%, and the financing approval time has been shortened from 5 days to 1 day;

[0187] 2. Based on transaction relationship feature vector Analysis shows that when the number of historical collaborations between a subcontractor and a general contractor is less than 3, the probability of a voucher dispute increases by 25%;

[0188] 3. Through cross-chain evidence storage, the owner can view the subcontractor's progress payment voucher in real time to avoid the risk of misappropriation of funds.

[0189] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A cross-chain evidence storage system for electronic accounting vouchers in supply chain finance, characterized by: include: Voucher data collection module, used to collect electronic accounting voucher data from the supply chain finance system; Multi-dimensional feature extraction module, used to extract the multi-dimensional feature vector of the electronic accounting document Blockchain alliance module, including multiple heterogeneous blockchain systems B={B1,B2,…,B m }; Cross-chain communication protocol module, which realizes cross-chain data transmission based on improved hash time lock contract (HTLC) and side chain technology; Distributed consensus verification module, using Byzantine fault-tolerant consensus algorithm in is the transaction set, is the set of verification nodes, α is the fault tolerance threshold; The voucher storage management module is used to retrieve, verify and update the stored electronic accounting vouchers.

2. The cross-chain evidence storage system for electronic accounting vouchers in supply chain finance according to claim 1 is characterized by: The multi-dimensional feature extraction module further includes: Time dimension feature extraction unit, extracting the voucher generation time t g , business occurrence time t b and the evidence storage time t s , and construct the time feature vector Amount dimension feature extraction unit, extracts voucher amount A, currency C and exchange rate E, and constructs amount feature vector Transaction relationship dimension feature extraction unit, extracting the identity information P1, P2 of both parties and the transaction type T t , and construct the transaction relationship feature vector Feature fusion unit, through the formula The multi-dimensional feature vectors are fused, where σ is the activation function, W is the weight matrix, and b is the bias vector.

3. The cross-chain evidence storage system for electronic accounting vouchers in supply chain finance according to claim 1 is characterized by: In the blockchain alliance module, each blockchain system B i With different block structures S i 、Consensus Mechanism M i and smart contract engine E i , and meet the heterogeneous compatibility conditions: where φ ij is the block structure conversion function, and satisfies: H(φ ij (x))=H(x)mod p Here H is a hash function and p is a secure prime number.

4. The cross-chain evidence storage system for electronic accounting vouchers in supply chain finance according to claim 1 is characterized by: The cross-chain communication protocol module adopts a cross-chain data verification mechanism based on zero-knowledge proof, specifically including: Data commitment generation unit, generates Merkle tree commitment C of electronic accounting document data D D =MerkleTree(D); Zero-knowledge proof generation unit generates proof π, which enables the verifier to verify C without revealing the specific content of D D effectiveness; The cross-chain data transfer unit is calculated by the formula: Transfer(C D ,π,B i ,B j )=Verify(π,C D )∧Relay(C D ,B j ) Realize from blockchain B i To blockchain B j , where Verify is the verification function and Relay is the relay function.

5. The cross-chain evidence storage system for electronic accounting vouchers in supply chain finance according to claim 1 is characterized by: The Byzantine fault-tolerant consensus algorithm adopted by the distributed consensus verification module The following steps are involved: Pre-preparation phase: the master node broadcasts the pre-preparation message Where v is the view number, n is the sequence number, and d is Abstract; Preparation phase: Verification node Send Prepare Message<PREPARE,v,n,d,i> , when receiving at least 2f+1 different verification nodes’ prepare messages, the verification node enters the commit phase, where Commit phase: Verification node sends commit message<COMMIT,v,n,d,i> , when receiving commit messages from at least 2f+1 different verification nodes, the verification node executes the transaction and updates the status; The algorithm meets the security conditions: and the active condition:

6. The cross-chain evidence storage system for electronic accounting vouchers in supply chain finance according to claim 1 is characterized by: The credential storage management module further includes: The evidence index construction unit builds an evidence index structure based on Bloom filter I = BloomFilter({h(D1),h(D2),…,h(D k )}), where D i is the i-th electronic accounting document, h is the hash function; Evidence retrieval unit, through the formula: Achieve fast retrieval of stored evidence, where q is the retrieval query; Evidence verification unit, through the formula: Verify(D,proof)=H(D)=extractRoot(proof) Verify the integrity of the proof, where proof is the proof of the proof and extractRoot is the function that extracts the root hash from the proof; The evidence update unit uses the version control mechanism to generate a new evidence for the updated certificate D′<D′,v+1,timestamp,signature> , and establish an association with the original evidence.

7. The cross-chain evidence storage system for electronic accounting vouchers in supply chain finance according to claim 1 is characterized by: Also includes: The data security enhancement module uses the homomorphic encryption algorithm E and the secure multi-party computing protocol to achieve: E(f(x1,x2,…,x n ))=f(E(x1),E(x2),…,E(x n )) Where f is any polynomial time computable function; The privacy protection module uses the zero-knowledge proof protocol Π = (Prover, Verifier) ​​to enable the verifier to verify the validity of the statement stmt without leaking any additional information: where ≈ c Denotes computational indistinguishability.

8. The cross-chain evidence storage system for electronic accounting vouchers in supply chain finance according to claim 1 is characterized by: The electronic accounting document has a multi-dimensional time stamp structure TS=[t1, t2, ..., t k ],in: t1 is the timestamp of the transaction, satisfying t1 = Hash(D||nonce1); t2 is the timestamp for generating the credential, satisfying t2 = Hash(t1||D′||nonce2); t3 is the timestamp of the first proof storage, satisfying t3 = Hash(t2||B1||nonce3); t i The timestamp of the i-th cross-chain evidence storage, satisfying t i =Hash(t i-1 ||B i-1 ||B i ||nonce i ); The timestamp sequence satisfies the chain verification relationship: The context i Context information generated for the i-th timestamp.

9. The cross-chain evidence storage system for electronic accounting vouchers in supply chain finance according to claim 1 is characterized by: The system further includes a risk assessment module, which performs risk assessment on electronic accounting documents through the following steps: Construct the credential feature matrix X = [x ij ] n×m , where x ij represents the jth feature of the i-th voucher; Train the risk assessment model M so that: M(X)=σ(W L ·σ(W L-1 ·…·σ(W1·X+b1)…+b L-1 )+b L ) Where W l and b l are the weight matrix and bias vector of the lth layer respectively, and σ is the activation function; Calculating risk scores in is the feature weight vector; According to the risk score R and the preset threshold θ, the credentials are divided into high risk (R>θ), medium risk (R∈[θ / 2,θ]) and low risk (R<θ / 2) categories.

10. The cross-chain evidence storage system for electronic accounting vouchers in supply chain finance according to claim 1 is characterized by: The system also includes an audit tracking module, which implements audit tracking of the entire life cycle of electronic accounting documents through the following steps: Construct a credential event graph G = (V, E), where the node set V = {v1, v2, ..., v n } represents the credential state, and the edge set E={e1,e2,…,e m } indicates a state transition event; For each event e i =(v j ,v k ,t i ,a i )Record timestamp t i and operating subject a i ; Define the audit path as a node sequence P = [v1, v2, ..., v k ],satisfy Constructing Audit Proof π P , such that: The proof is based on zero-knowledge proof technology to ensure the privacy and security of the audit process.

Images