Financial performance data integration method based on block chain

Abstract

The invention relates to the technical field of data integration, provides a financial performance data integration method based on a block chain, and aims to solve the problem of on-chain analysis distortion caused by the fact that the block chain cannot avoid original data errors. According to the method, a verification network is constructed through multi-source data collection, data logic verification is achieved in combination with a relevance map, a dynamic weight mechanism is introduced to quantify data source reliability, and a multi-level data verification system is formed. Meanwhile, a hierarchical error correction process is designed, slight errors are automatically repaired, correction consensus is achieved for serious errors through a hierarchical consensus mechanism, and the full life cycle of data is recorded through a holographic tracing chain. According to the method, the accuracy and the fault-tolerant capability of the block chain financial data are improved, the reliability and the traceability of data integration are guaranteed, and the method is suitable for performance data management of multiple scenes such as supply chain finance.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of financial data integration, in particular to a financial performance data integration method based on a block chain. BACKGROUND

[0002] The block chain technology is widely used in the field of financial performance data integration due to its tamper-proofing property, which can ensure the integrity and consistency of the data on the chain and provide a data basis for financial analysis, risk control, etc.

[0003] However, the block chain can only guarantee the tamper-proofing property of the data after being chained, and cannot solve the problem of input error of the original data. In actual application, the original data may be incorrect due to human input error, sensor failure, system interface anomaly, etc. Once such incorrect data is chained, it will cause deviation in subsequent performance analysis, transaction triggering, etc. based on the data on the chain, and even cause serious consequences such as loss of funds and decision-making errors.

[0004] In the prior art, although there are multi-source data verification mechanisms, they are mostly simple numerical range comparisons and do not consider the logical association between data, making it difficult to identify implicit errors. The weight of the data source is fixed and cannot be dynamically adjusted according to its real-time reliability, affecting the accuracy of verification. The data error correction process lacks a hierarchical consensus mechanism, either relying on centralized node decision-making with trust risks, or being cumbersome and inefficient, making it difficult to meet the high requirements of the financial field for data accuracy and processing timeliness.

[0005] Therefore, a financial performance data integration method based on a block chain is proposed, which intelligently identifies original data errors, dynamically optimizes verification logic, and efficiently implements consensus error correction, to solve the above technical problems. SUMMARY

[0006] Technical problems solved In the prior art, although there are multi-source data verification mechanisms, they are mostly simple numerical range comparisons and do not consider the logical association between data, making it difficult to identify implicit errors. The weight of the data source is fixed and cannot be dynamically adjusted according to its real-time reliability, affecting the accuracy of verification. The data error correction process lacks a hierarchical consensus mechanism, either relying on centralized node decision-making with trust risks, or being cumbersome and inefficient, making it difficult to meet the high requirements of the financial field for data accuracy and processing timeliness.

[0007] Technical solutions To achieve the above purposes, the present application provides the following technical solutions: a financial performance data integration method based on a block chain, comprising the following steps: S1. Intelligent data collection and preprocessing: Collecting financial performance related data through multi-source devices, including Internet of Things sensors and decentralized oracle network interfaces; standardizing the format of the collected data; using a rule-based matching algorithm to extract key features from the preprocessed data and match them with the initial model of the preset correlation graph, marking potential logical conflict data; encrypting the processed data packets and transmitting them to the temporary storage area of the blockchain node through the encrypted interface; S2. Dynamic verification and on-chain writing: Building and updating the financial data correlation graph, which contains entity nodes, attribute nodes and the correlation between nodes, and updating the correlation strength every week by algorithm from the on-chain data; smart contract analyzes the temporary storage area data, traverses the correlation rules in the correlation graph for logical verification; calculating the real-time score of each data source; S3. Intelligent error correction and process linkage: using the deviation rate formula Classify the error data, The deviation rate is The error data value is The correct data value is; for minor errors, automatically repair through the preset rule library, and for serious errors, start the hierarchical consensus correction process, and the consensus total score formula is , , , , , , The agreement rate of the corresponding level is When detecting data anomalies, trigger the corresponding early warning mechanism according to the error level, notify the relevant personnel and track the processing status;

[0008] Preferably, the Internet of Things sensor in step S1 includes: a GPS module with a positioning error of ≤5 meters and a sampling frequency of 1 / 30 seconds; a temperature and humidity sensor with a measurement range of -40℃~125℃ and an accuracy of ±0.3℃, a humidity measurement range of 0%~100%RH and an accuracy of ±2%RH; a weight sensor with a measurement range of 0~5000kg and an accuracy of 0.05%FS; an infrared counter with a sensing distance of 0~5 meters and a response time of ≤10ms.

[0009] Preferably, in step S1, the multi-source devices need to be authenticated by the blockchain node, and a unique public / private key pair is generated using the ECC algorithm. The private key is stored in the device's security chip, and the public key is written into the blockchain device whitelist. Device data that fails authentication will be rejected.

[0010] Preferably, in step S1, the processed data packet is encrypted using the AES-256 algorithm. The key is dynamically generated through the Diffie-Hellman key exchange protocol and updated once per hour. The encrypted data packet contains digest information based on the SHA-256 algorithm.

[0011] Preferably, the initial association rules of the association graph in step S2 are extracted from historical data using the Apriori algorithm, and the extraction formula is as follows: , For the confidence level of the association rule, For support, To increase the degree, when It was incorporated into the initial map at that time.

[0012] Preferably, the scoring in step S2 includes historical accuracy, response speed, and stability, wherein: the formula for historical accuracy is... , For historical accuracy, This represents the amount of valid data from the past 30 days. This represents the total data volume over the past 30 days; the response speed formula is... , For response speed, This is the actual delay time. The industry average latency is given; the stability formula is... , For stability, For continuous trouble-free operation time, The total duration is 30 days; a comprehensive score is calculated based on the weights of each data source and real-time ratings. , For the first The weight of each data source, For the first The score of each data source, when At that time, the data is written to the main blockchain chain; the data source score is determined by the formula. The calculations are performed, where 0.5, 0.3, and 0.2 are the weights for historical accuracy, response speed, and stability, respectively.

[0013] Preferably, the error classification threshold in step S3 is: the minor error threshold is... This could be a formatting error; the critical error threshold is... Or involving an amount ≥ 500,000 yuan; for In cases where the amount is less than 500,000 yuan, the decision is made manually by two professionals.

[0014] Preferably, the warning level in step S3 is divided into three levels: Level 1 warning is... For amounts ≥ 1 million RMB, notify the department director within 15 minutes, activate the emergency response team within 1 hour, and complete the handling within 4 hours; Level II warning is... For amounts between 500,000 and 1,000,000 yuan, notify the risk control manager within 30 minutes and complete the processing within 3 hours; Level 3 warning is... For amounts less than 500,000 yuan, the responsible person will be notified within 2 hours and the processing will be completed within 6 hours.

[0015] Preferably, the information recorded in the acquisition phase in step S4 includes the device number, device model, acquisition time, original data value, data encryption method, and original data hash value; the access log records the data access user ID, access time, access content hash, and access permission level.

[0016] Preferably, the permission verification process for the visual traceability query in step S4 is as follows: the queryer logs into the system through a CA certificate, the smart contract verifies the validity of the certificate and the permission level, and returns the query result corresponding to the permission after the verification is successful.

[0017] Beneficial effects Compared with existing technologies, this invention provides a blockchain-based method for integrating financial performance data, which has the following advantages: 1. This solution overcomes the limitations of single-dimensional data verification in existing technologies by constructing a data correlation graph. It can uncover implicit logical relationships between multi-source data, identify data anomalies from an overall logical perspective, and solve the problem of missed detections caused by neglecting data correlation in traditional verification methods. Simultaneously, the dynamic weight adjustment mechanism can flexibly adjust the weight of data sources in verification based on their real-time performance (such as stability and accuracy), overcoming the shortcomings of fixed-weight mode in adapting to changes in data source reliability. This makes the verification results more closely reflect the dynamic changes in actual business scenarios.

[0018] 2. This solution employs a tiered consensus mechanism, allocating voting weights based on the expertise and responsibilities of participating parties. This ensures compliance oversight at the core level, technical support at the professional level, and feedback from business practices at the general level, making error correction decisions more professional and authoritative, and reducing inefficiencies or errors caused by non-professional involvement. The intelligent early warning and automatic repair functions automate the handling of minor errors, avoiding the cumbersome steps of traditional manual error correction processes and significantly improving efficiency. Simultaneously, clear error classification standards ensure that serious errors are handled prudently.

[0019] 3. This solution establishes a holographic traceability chain to fully record the entire process of data collection, verification, error correction, and access, including the actions of participants at each stage, details of data changes, and related credentials. Compared to existing traceability methods that only record key nodes, this solution achieves complete transparency and traceability of data flow. Combined with device authentication, encrypted transmission, and access control, it effectively prevents data tampering and unauthorized access, providing complete evidence for auditing and supervision, and solving the problems of incomplete traceability information and weak security in traditional blockchain systems.

[0020] 4. This solution embeds financial business logic into the data verification process through a correlation graph, creating a closed loop between technical means and business rules. This overcomes the problem of disconnect between technical implementation and business needs in existing technologies. The hierarchical processing mechanism and early warning linkage mechanism in the dynamic error correction process can flexibly adapt processing strategies according to the severity of errors and the scope of business impact, ensuring that the technical solution can efficiently respond to the actual needs of different business scenarios and improving the practicality and applicability of blockchain technology in financial performance data integration. Attached Figure Description

[0021] Figure 1 This is a schematic diagram of step one of the present invention; Figure 2 This is a schematic diagram of step two of the present invention; Figure 3 This is a schematic diagram of step three of the present invention; Figure 4 This is a schematic diagram of step four of the present invention. Detailed Implementation

[0022] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0023] Please refer to 1~ Figure 4 This invention proposes a blockchain-based method for integrating financial performance data, comprising the following steps: Step 1: Intelligent Data Acquisition and Preprocessing (I) Deployment of Multi-Source Data Access Equipment Selection and Parameters: The logistics sensor uses a high-precision GPS module with a positioning error ≤5 meters, a sampling frequency of 1 time / 30 seconds, and supports BeiDou / GPS dual-mode positioning. The temperature and humidity sensor uses the SHT35 model, with a temperature measurement range of -40℃ to 125℃ and an accuracy of ±0.3℃, a humidity measurement range of 0% to 100%RH and an accuracy of ±2%RH, and a data transmission interface of RS485. The warehouse weight sensor uses the HBMZ6 series, with a measurement range of 0~5000kg, an accuracy of 0.05%FS, and is equipped with an overload protection device. The infrared counter uses a diffuse reflection type, with a sensing distance of 0~5 meters, a response time ≤10ms, and can identify objects as small as 5×5cm.

[0024] DON Interface Development: An interface based on Chainlink's Off-chain Reporting (OCR) protocol is developed, with a data update frequency set at 5 minutes per update. The interface uses TLS 1.3 encryption for transmission to ensure data integrity during transmission. When integrating with the customs declaration system, authentication is performed via API keys, accessing only fields relevant to the current business, such as customs declaration numbers, commodity codes, and declared weights. When integrating with third-party credit reporting platforms, a data access whitelist is set up, allowing only queries of necessary information such as credit scores and historical default records of specific companies.

[0025] Device authentication and fault tolerance: All connected devices (sensors, DON nodes) must be authenticated by the blockchain node. A unique public / private key pair is generated using the ECC algorithm. The private key is stored in the device's security chip, and the public key is written to the blockchain device whitelist. If device authentication fails (e.g., the public key is not on the whitelist), the data is rejected. When a single device is offline for ≤30 minutes, local caching (with a cache capacity ≥2 hours of data) is automatically activated. After reconnection, the data is synchronized to the blockchain node to prevent data loss.

[0026] Deployment Example: A batch of electronic components is being transported from Shenzhen to Shanghai. Three temperature and humidity sensors (located on the top, middle, and bottom layers of the cargo, respectively), one GPS module (installed in the driver's cab), and one weight sensor (installed at the bottom of the truck bed) are deployed. The DON interface simultaneously connects to the Shenzhen Customs declaration system and the credit information platform of a Shanghai warehousing company to obtain real-time customs declaration data and the warehousing company's credit information for this batch of goods. All devices are authenticated, and the local cache is set to generate a data snapshot every 10 minutes.

[0027] (ii) Edge node preprocessing Format standardization rules: Timestamp conversion uses Python's datetime module to uniformly convert non-standard times from various data sources (such as "2024-7-17 14:30" and "17 / 07 / 2024 14:30") into the format "YYYY-MM-DDTHH:MM:SS+HH:00"; numerical unit conversion is achieved through preset conversion factors, such as 1kg=0.001t, 1m=100cm, etc., and 6 decimal places are retained during the conversion process to reduce precision loss.

[0028] Encrypted data transmission: Preprocessed data packets are encrypted using the AES-256 algorithm. The key is dynamically generated by the edge nodes and blockchain nodes through the Diffie-Hellman key exchange protocol and is updated hourly. The encrypted data packets contain a data digest (SHA-256 hash). After receiving the data, the blockchain nodes verify the consistency of the digest to prevent tampering during transmission.

[0029] Preliminary Association Mapping Algorithm: A rule-based matching algorithm is used to extract key features (such as cargo ID, weight, and dimensions) from the preprocessed data packet and match them with the initial model of the association graph. For example, when "cargo weight 3.2t" is extracted, the algorithm automatically searches for related nodes such as "vehicle approved load capacity" and "maximum load capacity of storage location" in the graph. If there is a case where "3.2t > 4t" (storage location load capacity), it is immediately marked as a "potential logical conflict," and the ID and specific value of the conflicting node are recorded.

[0030] Data cleaning example: A sensor uploads data as "temperature 25.6℃, humidity 65%, time 2024-7-17 14:30". The edge node first converts the time to "2024-07-17T14:30:00+08:00", finds that the humidity of 65% exceeds the preset reasonable range (40%~60%), marks it as "humidity abnormal", encrypts it and sends it to the blockchain temporary storage area.

[0031] Step 2: Dynamic Verification and On-Chain Writing (I) Real-time analysis of correlation maps Initial graph construction and update mechanism: Initial association rules are extracted from historical data using the Apriori algorithm, with the following formula: ,in For the confidence level of the association rule, Support (frequency of rule occurrence) To improve the reliability of the rules. When The initial graph is added when the transaction is linked to the order amount (e.g., "order amount - invoice amount"). The graph is automatically updated weekly via a Python script, and transactions are automatically added to the graph when 100 consecutive transactions satisfy a new link.

[0032] Logical verification process: The smart contract first parses the key information of the data packet in the temporary storage area, and then traverses the relevant association rules in the graph, verifying them one by one. Taking a certain accounts receivable data as an example, the verification process is as follows: Extract the "order amount 100,000 yuan", "invoice amount 113,000 yuan", and "VAT rate 13%" from the data packet. Verify that "invoice amount = order amount × (1 + VAT rate)", and 10 × 1.13 = 113,000 yuan. The rule is satisfied. Extract "Payment period 30 days" and "Contractual agreed days 30 days", and verify that "Payment period ≤ Contractual agreed days" meets the rule; Extract "Logistics receipt time 2024-07-10" and "Invoice issuance time 2024-07-12", verify "Logistics receipt time ≤ Invoice issuance time", the rule is satisfied, and finally it is judged as "Logic passed".

[0033] Example of conflict handling: If the invoice amount is 120,000 yuan, the order amount is 100,000 yuan, and the VAT rate is 13%, then the invoice amount is greater than the order amount by 1.13. The logic check fails, the conflict point is recorded, and the data is entered into the exception pool.

[0034] (II) Dynamic weight calculation and comprehensive judgment Data source scoring details: Historical accuracy: The formula is as follows ,in Indicates historical accuracy. This represents the amount of valid data from the past 30 days. This represents the total amount of data transmitted in the past 30 days. For example, a sensor transmitted 1000 data points in 30 days, of which 5 were incorrect. , ,but .

[0035] Response speed: The formula is ,in Indicates response speed. The actual data transmission delay time (in seconds). This is the industry average latency (taken as 2 seconds). If the actual latency is 1 second, then... .

[0036] Stability: The formula is ,in Indicates stability. Continuous trouble-free operation time (in hours). The total duration is 30 days (720 hours). If a failure occurs and lasts for 1 hour, then... , .

[0037] Data source score: Formula is ,in , , They are respectively , , The weight. For example, a data source. , , ,but .

[0038] Example of overall score calculation: The formula for the overall score is as follows ,in For the first The weight of each data source, For the first The scores from each data source. The weights and scores of the three data sources for a given data point are as follows: Logistics Sensors ( , ), warehouse data ( , ), DON data ( , ),but If the value is ≥0.6, the data is considered valid and written to the main blockchain.

[0039] Step 3: Intelligent Error Correction and Process Linkage (I) Classification Error Identification and Handling Error grading threshold: Minor errors are defined as follows: Or it may be a formatting error (such as a misplaced decimal point), in which ( The deviation rate, This is an incorrect data value. (Correct data value); a serious error is defined as... Or the amount involved is ≥ 500,000 yuan; Furthermore, if the amount is less than 500,000 yuan, it shall be manually determined by two professional personnel.

[0040] The automatic error correction rule base includes rules for minor errors such as formatting error correction (e.g., adding decimal points, correcting units) and calculation error correction (e.g., recalculating VAT, discount amounts). For example, "Accounts receivable 1234 yuan" is an incorrect value. The correct value is 12340 yuan. ),but However, because it is a formatting error (misaligned decimal point), it is still classified as a minor error.

[0041] Critical Error Layered Consensus Process: Proposal initiation: The warehouse administrator logs into the web interface, enters the error data ID (e.g., "DATA-20240717001"), fills in the correction value (e.g., the weight is corrected from 3.5t to 3.2t), uploads the video and photos of the re-weighing as supporting materials, and submits the proposal.

[0042] Core layer review: The central bank regulatory node and industry association receive the proposal within 2 hours, and review the authenticity and compliance of the supporting materials. If one of the two core nodes agrees, the proposal will proceed to the professional layer for voting.

[0043] Multi-level voting: The formula for the total consensus score is as follows ,in (Core layer weight) (Professional level weight) (Ordinary layer weights); , , The approval rates are for the core layer, professional layer, and general layer, respectively. Let's assume the approval rate for the core layer is... The professional panel voted 2 in favor and 1 against. ), 3 votes in favor and 2 votes against in the ordinary tier ( ),but The proposal was deemed approved.

[0044] Example of correction execution: The smart contract corrects the data from 3.5t to 3.2t, writes it to the main chain, and sends a "pause payment" instruction to the payment system until the correction is completed and a "resume payment" instruction is sent.

[0045] (II) Triggering of the Early Warning and Linkage Mechanism Warning levels and response procedures: Level 1 warning ( For cases involving amounts ≥ 1 million RMB: Notify the department director within 15 minutes, activate the emergency response team within 1 hour, and complete the handling within 4 hours.

[0046] Level II warning ( For amounts between 500,000 and 1,000,000 yuan: notify the risk control manager within 30 minutes and complete the processing within 3 hours.

[0047] Level III warning ( For amounts less than 500,000 yuan: the person in charge will be notified within 2 hours and the processing will be completed within 6 hours.

[0048] The abnormal notification content includes: "[Warning] Sensor A123 data is abnormal, business order number: accounts receivable 20240717001, conflict point: weight 3.5t > load capacity of the storage location 3t, processing time limit: 6 hours"; the system work order includes the warning level (red), associated data ID, conflict details, and processing suggestions (such as "please re-weigh and upload supporting documents").

[0049] Processing tracking mechanism: The status of alert processing (unprocessed, in progress, completed) is recorded via blockchain and updated every 2 hours. If the alert is not processed within the processing time limit, an escalation notification is automatically sent to a higher-level person (such as the department director), and a red countdown reminder is displayed on the system homepage.

[0050] Step Four: Holographic Traceability and Audit Support (a) Data record fields throughout the entire lifecycle Data Acquisition Phase: Device ID, Device Model, Acquisition Time (accurate to milliseconds), Raw Data Value, Data Encryption Method (e.g., AES-256), Raw Data Hash Value.

[0051] Verification phase: weight values ​​of each data source, rule IDs that passed the graph verification, rule IDs that failed (if any), smart contract call records (such as call time and return results), and comprehensive score.

[0052] Error correction phase (if any): Proposer ID, proposal time, hash value of supporting materials, voting results of each voting node (agree / disagree / abstain), voting time, data values ​​before and after correction, correction time, and smart contract address for executing the correction.

[0053] Access logs record the data access user ID, access time, access content hash, and access permission level to ensure that "who accessed, when accessed, and what was accessed" is traceable.

[0054] Invocation phase: Queries ID, query time, query purpose (e.g., audit, risk control), and query result hash value.

[0055] (ii) Visual traceability query function Query permission verification process: The queryer logs into the system with a CA certificate. The smart contract (traceability management contract) verifies the validity of the certificate and the permission level (e.g., auditor permission ≥ level 3, ordinary employee permission = level 1). After successful verification, the query results corresponding to the permission are returned.

[0056] Query entry: The web interface provides multiple query methods such as data ID, business order number, and device number, and supports fuzzy search (e.g., enter "20240717" to query all data for the day).

[0057] Display content: The entire lifecycle of the data is displayed in a timeline format, and each node can be clicked to view detailed information (such as clicking "Collection Stage" to view the encryption and decryption process of the original data); it provides data change comparison charts (such as bar charts of data values ​​before and after correction), pie charts of voting results, and other visual charts.

[0058] Access control: Different roles have different query permissions. Auditors can view all fields, while ordinary employees can only view non-sensitive information (such as data values ​​and processing status). Permission settings are managed through smart contracts (traceability management contracts), and modification of permissions requires the approval of a vote by the core layer nodes.

[0059] Finally, it should be noted that the above embodiments are merely examples for clearly illustrating the present invention and are not intended to limit the implementation. Those skilled in the art will recognize that other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively list all possible implementations. However, obvious variations or modifications derived therefrom are still within the scope of protection of this invention.

Claims

1.A blockchain-based financial performance data integration method, characterized in that: Comprise the following steps: S1. Intelligent data acquisition and preprocessing: Collecting financial performance related data through multi-source devices, including Internet of Things sensors and decentralized oracle network interface; Format standardization processing of collected data; Using rule-based matching algorithm, extract key features from preprocessed data and match with preset correlation graph initial model, mark potential logical conflict data; The data packet processed is encrypted, and transmitted to the temporary storage area of the block chain node through the encrypted interface; S2. Dynamic verification and on-chain writing: Build and update financial data correlation graph, the graph contains entity nodes, attribute nodes and the correlation between nodes, new correlation rules are mined from the chain data every week and the correlation strength is updated; Smart contract analyzes temporary storage area data, traverses correlation rules in correlation graph for logical verification; Calculate the real-time score of each data source; S3. Intelligent error correction and process linkage: Adopting deviation rate formula Classifying error data, For deviation rate, For error data value, For correct data value; Minor errors are automatically repaired by a preset rule library, and serious errors start a layered consensus correction process. The total consensus score formula is , , , are the weights of the core layer, professional layer and general layer respectively, , , are the agreement rates of the corresponding levels respectively; when data anomalies are detected, the corresponding warning mechanism is triggered according to the error level, relevant personnel are notified and the processing status is tracked. S4. Holographic traceability and audit support: Record the whole life cycle information of data, including collection stage, verification stage, error correction stage, access log and related data of calling stage; Provide visual traceability query function, different roles of users can view corresponding data information according to preset permissions, and permission modification needs to be voted and agreed by core layer nodes. 2.The blockchain-based financial performance data integration method of claim 1, wherein: The Internet of Things sensor in step S1 includes: GPS module with positioning error ≤ 5 meters and sampling frequency of 1 time / 30 seconds; Temperature measurement range-40℃~125℃, accuracy ±0.3℃, humidity measurement range 0%~100%RH, accuracy ±2%RH, temperature and humidity sensor; Measurement range 0~5000kg, accuracy 0.05%FS, weight sensor; Sensing distance 0~5 meters, response time ≤ 10 ms, infrared counter. 3.The blockchain-based financial performance data integration method of claim 1, wherein: The multi-source device in step S1 needs to be authenticated by the block chain node, and ECC algorithm is used to generate a unique public key / private key pair for the device. The private key is stored in the device security chip, and the public key is written into the block chain device whitelist. The data of the device that fails the authentication will be rejected. 4.The blockchain-based financial performance data integration method of claim 1, wherein: The data packet processed in step S1 is encrypted by AES-256 algorithm, and the key is dynamically generated by Diffie-Hellman key exchange protocol, updated once an hour, and the encrypted data packet contains summary information based on SHA-256 algorithm. 5.The blockchain-based financial performance data integration method of claim 1, wherein: The initial association rules of the correlation graph in step S2 are extracted from the historical data by the Apriori algorithm, and the extraction formula is , is the confidence of the association rule, is the support, is the lift, and when is included in the initial graph. 6.The blockchain-based financial performance data integration method of claim 1, wherein: The scoring in step S2 includes historical accuracy, response speed, and stability, wherein: the formula for historical accuracy is... , For historical accuracy, This represents the amount of valid data from the past 30 days. This represents the total data volume over the past 30 days; the response speed formula is... , For response speed, This is the actual delay time. The industry average latency is given; the stability formula is... , For stability, For continuous trouble-free operation time, The total duration is 30 days; a comprehensive score is calculated based on the weights of each data source and real-time ratings. , For the first The weight of each data source, For the first The score of each data source, when At that time, the data is written to the main blockchain chain; the data source score is determined by the formula. The calculations are performed, where 0.5, 0.3, and 0.2 are the weights for historical accuracy, response speed, and stability, respectively. 7.The blockchain-based financial performance data integration method of claim 1, wherein: The error classification threshold in step S3 is: a minor error threshold of or belongs to a format error; Serious error threshold is or involving an amount of ≥ 500,000 yuan; for and an amount of < 500,000 yuan, determined by 2 professional staff manually. 8.The blockchain-based financial performance data integration method of claim 1, wherein: The pre-warning level in step S3 is divided into three levels: first level pre-warning is or involves an amount of ≥ 1 million yuan, informs the department director within 15 minutes, starts the emergency group within 1 hour, and completes the processing within 4 hours; second level pre-warning is or an amount of 50-100 million yuan, informs the risk control manager within 30 minutes, and completes the processing within 3 hours; third level pre-warning is and an amount of <50 million yuan, informs the person in charge within 2 hours, and completes the processing within 6 hours. 9.The blockchain-based financial performance data integration method of claim 1, wherein: The information recorded in the collection stage in step S4 includes device number, device model, collection time, original data value, data encryption method and original data hash value; The access log records data access person ID, access time, access content hash and access permission level. 10.The blockchain-based financial performance data integration method of claim 1, wherein: The permission verification process of visual traceability query in step S4 is: the inquirer logs in the system through the CA certificate, the smart contract verifies the validity of the certificate and the permission level, and returns the query result corresponding to the permission after verification.

Images