Enterprise green credit checking method and system based on block chain

Abstract

The invention discloses an enterprise green credit checking method and system based on a block chain, and particularly relates to the technical field of block chains, and the method comprises the steps: collecting green behavior data from data sources of a plurality of enterprises, summarizing the green behavior data to obtain initial data, and carrying out the authenticity and integrity verification to form an upper chain data set; in the uplink process, acquiring uplink information and analyzing an uplink time window to judge whether an early sign of insufficient data quality exists or not; if it is found that the quality is insufficient, feature extraction is carried out on an uplink time window, the efficiency and the antagonism of an encryption algorithm are analyzed, the time window is divided into a high-quality window or a low-quality window through fuzzy logic reasoning, and data of the low-quality window are uplink again; and finally, carrying out green credit scoring on the enterprise through high-quality uplink data. According to the method, potential hidden dangers in data transmission and encryption can be found in advance, window data with insufficient quality signs can be automatically relinked, and the consistency and reliability of system data are ensured.

Description

Technical Field

[0001] The present invention relates to the field of blockchain technology. More specifically, the present invention relates to a method and system for enterprise green credit verification based on blockchain. Background Art

[0002] With the increasing global attention to sustainable development and environmental protection, more and more enterprises are required to disclose their green behaviors and environmental protection measures so that regulatory agencies, investors, and consumers can evaluate the environmental performance of enterprises. Green credit verification has become an important means of evaluating enterprises' environmental responsibilities and can reflect the environmental protection behaviors of enterprises in aspects such as resource use, carbon emissions, and waste treatment. However, traditional green credit verification often relies on manual records and centralized database storage, suffering from problems such as data tampering and insufficient transparency, making it difficult to ensure the fairness and reliability of data.

[0003] As a distributed ledger technology, blockchain technology has become an effective solution for solving the problem of reliable data storage due to its characteristics such as decentralization, immutability, and transparency. Blockchain stores data through a distributed node network, ensuring the immutability and traceability of data. These characteristics make its application in green credit verification have great potential, which can improve the transparency and credibility of green behavior data, prevent data forgery and tampering, and provide a more reliable data basis for the green credit scoring of enterprises.

[0004] With the support of blockchain, a green credit verification system can upload the green behavior data of enterprises to the blockchain, forming an open and transparent data chain. During the uploading process, the data is protected through encryption algorithms, and at the same time, a multi-verification mechanism is used to ensure the integrity and authenticity of the data. Storing the green behavior data in the blockchain ensures the immutability of the data throughout its life cycle, enabling any data change or anomaly to be traced, thereby realizing the comprehensive supervision and credible evaluation of enterprises' green behaviors.

[0005] However, during the uploading process, due to possible fluctuations in the efficiency and security of data uploading, there may be early signs of insufficient data quality within some uploading time windows, making it difficult to guarantee the stability and consistency of the uploaded data. On the other hand, the green credit scoring model of enterprises needs to rely on accurate and complete green behavior data. Therefore, once the quality of the uploaded data is not high, it will directly affect the accuracy and fairness of the green credit scoring. Therefore, a method and system for enterprise green credit verification based on blockchain are proposed here to solve the above problems. Summary of the Invention

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] Enterprise green credit verification method based on blockchain, comprising the following steps:

[0008] Collect the respective green behavior data of enterprises from the data sources of multiple enterprises to obtain multiple green behavior data sets, and then summarize them to obtain an initial data set;

[0009] Perform preset verification operations on the initial data set to ensure the authenticity and integrity of the data, and obtain the verified data set for uploading to the blockchain;

[0010] Perform an operation of uploading the data set for uploading to the blockchain, and obtain information during the operation of uploading to the blockchain to obtain an information set of the uploading process and perform an analysis of the uploading time window. According to the result of the analysis of the uploading time window, determine whether there are early signs of insufficient quality of the data uploaded during the uploading time window;

[0011] When there are early signs of insufficient quality of the data uploaded to the blockchain, extract the characteristics of the uploading time window, then perform efficiency analysis and adversarial analysis on the encryption algorithms used respectively, perform fuzzy logic reasoning on the results of the efficiency analysis and the adversarial analysis together, and divide the uploading time window into a high-quality uploading window or a low-quality uploading window;

[0012] Perform a re-upload operation on the content uploaded in the data set for uploading to the blockchain corresponding to the low-quality uploading window;

[0013] When conducting enterprise green credit verification, obtain all the green behavior data of the corresponding enterprises from the data uploaded in all high-quality uploading windows, and then use a preset green credit scoring model to evaluate the green credit of the enterprises and generate corresponding green credit scores.

[0014] In a preferred embodiment, the verification operation includes confirmation of data sources, verification of data timeliness, and verification of data consistency.

[0015] In a preferred embodiment, when the green credit score is within a preset standard score range, the early warning mechanism is not triggered. When the green credit score is not within the preset standard score range, the early warning mechanism is triggered, and a green credit early warning signal is sent to the corresponding enterprise.

[0016] In a preferred embodiment, during the process of performing the operation of uploading the data set for uploading to the blockchain, there are multiple preset uploading time windows, and a random encryption algorithm is obtained for each uploading time window for uploading encryption.

[0017] In a preferred embodiment, the analysis of the uploading time window refers to:

[0018] Obtain the data volume of each piece of content uploaded within the upload time window, the timestamp at the start, and the timestamp at the end. Subtract the start timestamp from the end timestamp to obtain the upload time. Then calculate the upload efficiency value for each piece of uploaded content, where the upload efficiency value is obtained by dividing the data volume of the uploaded content by the upload time. Next, calculate the average value PJi and the standard deviation BJi of all upload efficiency values within the upload time window; i is the number of the upload time window.

[0019] In a preferred embodiment, when determining whether there are early signs of insufficient quality of uploaded data during the upload within the upload time window according to the analysis result of the upload time window, it refers to:

[0020] Obtain the average value PJi and the standard deviation BJi of all upload efficiency values within the upload time window, and the encryption algorithm randomly obtained for the upload time window i. Look up the corresponding preset standard upload interval 1 and standard upload interval 2 according to the type of the encryption algorithm. If it is satisfied that the average value PJi falls within the range of the standard upload interval 1 and the standard deviation BJi falls within the range of the standard upload interval 2, then generate a normal signal. If it is not satisfied that the average value PJi falls within the range of the standard upload interval 1 and the standard deviation BJi falls within the range of the standard upload interval 2, then generate an abnormal signal. When the abnormal signal is generated, it indicates that there are early signs of insufficient quality of uploaded data in the upload time window.

[0021] In a preferred embodiment, the content of feature extraction includes the efficiency information of the encryption algorithm within the upload time window and the adversarial information of the encryption algorithm within the upload time window;

[0022] In a preferred embodiment, performing efficiency analysis on the adopted encryption algorithms respectively refers to:

[0023] Obtain the total execution time of the encryption operation, the total number of bytes of the encrypted data, the preset security score and data block size corresponding to the type of the encryption algorithm, the average occupancy rate of system resources during the encryption process, and whether the type of the encryption algorithm supports parallel computing from the efficiency information;

[0024] Then perform the following calculations:

[0025] Tenc represents the total execution time of the encryption operation, Ddate represents the total number of bytes of the encrypted data, Ssec represents the preset security score corresponding to the encryption algorithm type, Renc represents the average occupancy rate of system resources during the encryption process, Penc represents the parallelization efficiency. When the encryption algorithm type supports parallel computing, Penc takes the value of 1. When the encryption algorithm type does not support parallel computing, Penc takes the value of 0. Bsize is the standard data block size, Bj is the data block size corresponding to the encryption algorithm type j, Aj is the preset adjustment coefficient corresponding to the encryption algorithm type j. There are a total of N types of encryption algorithms, and the sum of the preset adjustment coefficients corresponding to the N types of encryption algorithms is one. XLI represents the encryption efficiency index.

[0026] In a preferred embodiment, the adversarial analysis of the encryption algorithm used means:

[0027] Obtain the encryption intensity data corresponding to each preset encryption intensity index, the number of network attacks and self - protection times within the on - chain time window, as well as the intensity of each network attack and self - protection from the adversarial information;

[0028] Then perform the following calculations:

[0029] C is a constant, m is the number of the encryption intensity index, fm is the encryption intensity data corresponding to the encryption intensity index m, M is the total number of encryption intensity indices, p is the fluctuation number of the on - chain time window. Fluctuation refers to the security fluctuation caused by network attacks or self - protection. The total number of network attack times and self - protection times is P, kp represents the fluctuation intensity corresponding to the fluctuation number p of the on - chain time window. Fluctuation intensity refers to the network attack intensity or self - protection intensity, vp represents the fluctuation coefficient corresponding to the fluctuation number vp of the on - chain time window. The fluctuation coefficient is composed of the negative fluctuation proportion coefficient corresponding to network attacks and the positive fluctuation proportion coefficient corresponding to self - protection, and satisfies that the sum of all negative fluctuation proportion coefficients is negative one, and the sum of all positive fluctuation proportion coefficients is one. CRI represents the encryption confrontation index.

[0030] In a preferred embodiment, the logic of fuzzy logic reasoning is:

[0031] Take both the encryption confrontation index and the encryption efficiency index of the on - chain time window as input variables, take the division type of the on - chain time window as the output variable, perform fuzzy processing on the input variables, convert the values of the input variables into fuzzy sets, perform fuzzy processing on the output variable, convert the output variable into a fuzzy set, formulate fuzzy rules to describe the on - chain quality under different combinations of data types, and infer the division type of the on - chain time window by passing the fuzzy - processed input variables through the fuzzy rules.

[0032] In a preferred embodiment, a blockchain-based enterprise green credit verification system includes:

[0033] An enterprise data collection module that collects the respective green behavior data of enterprises from the data sources of multiple enterprises, obtains multiple green behavior data sets, and then aggregates them to obtain an initial data set;

[0034] A verification module that performs preset verification operations based on the initial data set to ensure the authenticity and integrity of the data, and obtains a verified data set for uploading to the chain;

[0035] An on-chain analysis module that performs an on-chain operation on the data set for uploading to the chain, and obtains information during the on-chain operation process to obtain an on-chain process information set and perform an on-chain time window analysis. According to the on-chain time window analysis result, it is judged whether there are early signs of insufficient quality of the data for uploading to the chain when uploading within the on-chain time window;

[0036] An encryption algorithm analysis module, when there are early signs of insufficient quality of the data for uploading to the chain, extracts features from the on-chain time window, then performs efficiency analysis and adversarial analysis on the encryption algorithms used respectively, performs fuzzy logic reasoning on the results of the efficiency analysis and adversarial analysis together, and divides the on-chain time window into a high-quality on-chain window or a low-quality on-chain window;

[0037] An optimization module that performs a re-on-chain operation on the content for uploading to the chain in the data set for uploading to the chain corresponding to the low-quality on-chain window;

[0038] A credit verification module, when performing enterprise green credit verification, obtains all the green behavior data of the corresponding enterprise from the data for uploading to the chain in all high-quality on-chain windows, then uses a preset green credit scoring model to evaluate the green credit of the enterprise, generates a corresponding green credit score, and when triggering the early warning mechanism, sends a green credit early warning signal to the corresponding enterprise.

[0039] The technical effects and advantages of the present invention:

[0040] Through the decentralized and immutable characteristics of blockchain technology, the present invention stores the green behavior data of enterprises on the chain, effectively preventing the risk of malicious data tampering. Encryption algorithms are used to protect the data during the on-chain process to ensure the security of data transmission and storage, and avoid the distortion of credit scores caused by data tampering. By combining distributed storage and encryption technology, the integrity and authenticity of the data are ensured, providing a reliable data basis for green credit scoring.

[0041] The present invention sets multiple data uploading time windows and randomly selects encryption algorithms within each window for data uploading. By analyzing indicators such as the data volume, timestamp, and encryption efficiency within the data uploading time windows, dynamic monitoring of the quality of the uploaded data is achieved. Through feature extraction and encryption efficiency analysis, early quality problems in the data uploading process can be identified, and potential hidden dangers in data transmission and encryption can be discovered in advance. For the window data showing signs of insufficient quality, the present invention will automatically re-upload it to ensure the consistency and reliability of the system data.

[0042] The green credit score of the present invention is based on the green behavior data in high-quality data uploading windows, effectively avoiding the interference of low-quality data on the scoring results. By screening and using high-quality data, the accuracy of the green credit score is improved. The scoring model uses a convolutional neural network to analyze and calculate the green behavior data, and utilizes big data and artificial intelligence technologies to comprehensively evaluate the environmental protection performance of enterprises, ensuring the fairness and authority of the scoring results and providing a more scientific basis for enterprise credit verification.

[0043] Through the open and transparent mechanism of the blockchain, the present invention uploads the green behavior data of enterprises, making the data transparently visible to stakeholders such as regulatory agencies, consumers, and investors, increasing the credibility of the enterprises' environmental protection performance. The green credit score of the enterprises is traceable, and any modification or abnormality of the data can be traced, enhancing the transparency of the credit verification system and further promoting the enthusiasm of enterprises to fulfill their environmental responsibilities. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] For the convenience of those skilled in the art to understand, the present invention will be further described below in conjunction with the accompanying drawings;

[0045] Figure 1 is the schematic diagram of the method for enterprise green credit verification based on blockchain in the present invention.

[0046] Figure 2 is the schematic diagram of the system for enterprise green credit verification based on blockchain in the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0047] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0048] Refer to Figure 1 - Figure 2 to obtain the following embodiments:

[0049] Example 1: Blockchain is a decentralized distributed ledger technology that ensures the security and immutability of data through encryption. Blockchain stores data in chunks, with each block connected to the previous one through encryption to form a chain. Every transaction is recorded in a block, and once the data is recorded, it cannot be changed, thus ensuring the transparency and reliability of the data.

[0050] Enterprise green credit refers to the fulfillment of an enterprise's social responsibilities in aspects such as environmental protection, resource conservation, and pollution control. This includes the impact of the enterprise on the environment during the production process, whether it complies with environmental protection regulations, and whether it has taken green innovation and energy-saving measures. The level of an enterprise's green credit directly affects its market competitiveness and brand image.

[0051] Verification refers to the process of reviewing and validating an enterprise's green actions, environmental protection measures, compliance, etc. Verification can be carried out by third-party institutions, regulatory authorities, or self-auditing, etc., with the aim of ensuring that the enterprise truly fulfills its environmental protection responsibilities and meets the standards of green credit assessment.

[0052] The transparency and immutability of blockchain can ensure the authenticity of an enterprise's green credit verification data. When an enterprise conducts green projects or environmental protection measures, relevant data (such as carbon emissions, resource usage, green certifications, etc.) can be recorded on the blockchain in real time. Due to the immutability of blockchain data, third-party institutions and regulatory authorities can verify an enterprise's green behavior through the blockchain without relying on traditional and easily manipulable auditing and reporting mechanisms.

[0053] As environmental protection requirements become increasingly strict, an enterprise's green credit has become an important criterion for measuring its fulfillment of social responsibilities. Through blockchain, an enterprise's green behavior is recorded and verified publicly and transparently, which not only helps improve the enterprise's reputation among consumers and investors but also increases its opportunities to obtain policies such as green financial support and tax cuts.

[0054] Application scenarios: Green finance: Banks and financial institutions can decide whether to provide green loans, low-interest financing, or green investments based on an enterprise's green credit record (verified through blockchain).

[0055] Environmental protection compliance: Governments and regulatory authorities can use blockchain technology to continuously monitor and review an enterprise's environmental protection compliance to ensure that the enterprise complies with environmental protection laws and regulations.

[0056] Supply chain transparency: Enterprises can record information such as the procurement of green production raw materials, production process, and waste management on the blockchain to ensure the environmental protection compliance of the supply chain and enhance supply chain transparency.

[0057] Green Certification: Third-party certification agencies can use blockchain technology to record and verify the green certification process of enterprises, reducing cheating and data tampering problems in the certification process.

[0058] Based on the above content, the present invention proposes a method for verifying the green credit of enterprises based on blockchain, including the following steps:

[0059] Collect the green behavior data of each enterprise from the data sources of multiple enterprises to obtain multiple green behavior data sets, and then summarize them to obtain an initial data set; the core of this step is to comprehensively and widely collect the green behavior data of enterprises, covering various indicators such as environmental protection activities, carbon footprint, energy consumption, and waste treatment of each enterprise. By collecting and summarizing the green behavior data from multiple data sources, a unified data set can be established, providing a complete and rich basis for subsequent analysis. This multi-data-source collection method can ensure that the green behavior of enterprises is recorded in multiple dimensions, avoiding one-sided or distorted data, and laying a foundation for data preparation before uploading to the blockchain. By collecting a comprehensive and detailed data set, not only can the green performance of individual enterprises be reflected, but also comparisons between enterprises can be realized, thereby enhancing the evaluation breadth and depth of the entire system.

[0060] Perform preset verification operations on the initial data set to ensure the authenticity and integrity of the data, and obtain the verified data set for uploading to the blockchain; to ensure the fairness and authority of the green credit verification, it is crucial to verify the authenticity and integrity of the data. This step aims to eliminate potentially false, inaccurate, or inconsistent data through means such as data source verification, timeliness check, and consistency comparison. This operation ensures the reliability of the data, making the data uploaded to the blockchain not only have clear sources and timely updates, but also have a unified format and standard, avoiding misjudgment due to inconsistent data. The verified data set reflects the true green behavior of enterprises, providing high-quality data input for subsequent uploading to the blockchain and analysis, and ensuring the scientific nature and credibility of the entire credit verification process.

[0061] Perform the operation of uploading the data collection to the blockchain, obtain information during the uploading operation process, get the information set of the uploading process and conduct an analysis of the uploading time window. Determine whether there are early signs of insufficient quality of the uploaded data when uploading within the uploading time window according to the analysis results of the uploading time window. The uploading operation not only permanently records the data in the blockchain, but also records the information during the uploading process (such as uploading time, data size, time used, etc.) for further quality analysis. By monitoring the uploading process and analyzing the time window, potential problems in the uploading process, especially early signs of insufficient data quality, can be identified. For example, if the uploading time used is too long within a certain time window, it may indicate a bottleneck in the encryption or transmission process. This step discovers quality problems in the uploading process early by analyzing the uploading efficiency of each time window, takes measures in advance to ensure the stability and security of the uploaded data, thereby improving the reliability of the overall system.

[0062] When there are early signs of insufficient quality of the uploaded data, extract features from the uploading time window, and then conduct efficiency analysis and adversarial analysis on the encryption algorithms used respectively. Combine the results of the efficiency analysis and adversarial analysis for fuzzy logic reasoning, and divide the uploading time window into a high-quality uploading window or a low-quality uploading window. When early signs of insufficient quality of the uploaded data are found, through further feature extraction and in-depth analysis of the encryption algorithm, the root cause of the problem can be identified more accurately. Efficiency analysis evaluates the operation of the encryption algorithm within the time window, such as encryption speed and resource usage rate; adversarial analysis detects the security and anti-attack ability of the encryption algorithm. By comprehensively evaluating these analysis results through fuzzy logic reasoning, the uploading time window can be flexibly divided into high-quality or low-quality. The goal of this step is to conduct quality classification in a more intelligent way, provide an opportunity to correct the low-quality uploading window, ensure the stability of high-quality data at the same time, and can concentrate resources on optimizing the low-quality uploading window to optimize the overall data storage quality of the blockchain.

[0063] Perform the re-upload operation on the uploaded content in the data collection corresponding to the low-quality uploading window; low-quality uploaded data may affect the accuracy of the enterprise's green credit verification, so a re-upload operation is required. By re-uploading, it can ensure that the data meets the required quality standards, especially after re-selecting the encryption algorithm and adjusting the encryption method, improving the integrity and consistency of the data. This step helps the system automatically correct abnormal situations that may exist in the uploading process and reduce the impact of data defects on the enterprise credit assessment.

[0064] When conducting enterprise green credit verification, all green behavior data of the corresponding enterprise is obtained from the data uploaded to the high-quality upload windows, and then the preset green credit scoring model is used to evaluate the enterprise's green credit and generate the corresponding green credit score. The final green credit verification process ensures that data is only extracted from high-quality upload windows, eliminating the interference of low-quality data and ensuring the true and accurate reflection of the enterprise's green behavior. Through the preset green credit scoring model, a comprehensive and systematic evaluation of the enterprise's environmental protection performance can be carried out. The scoring results can not only provide a reference value for the enterprise's green credit but also serve as a standard for regulatory authorities, consumers, and partners to measure the enterprise's sustainable development. If the enterprise's green credit score does not meet the preset standard, an early warning mechanism can be triggered to prompt the enterprise to improve its green behavior, playing a role in motivating the enterprise to be more environmentally friendly and promoting the sustainable development of the entire industry.

[0065] The verification operations include but are not limited to data source confirmation, data timeliness verification, and data consistency check, aiming to ensure that the verified data set reflects the enterprise's true green behavior. The verification techniques are well-known existing technologies in this field and will not be elaborated here.

[0066] When the green credit score is within the preset standard score range, the early warning mechanism is not triggered. When the green credit score is not within the preset standard score range, the early warning mechanism is triggered to send a green credit early warning signal to the corresponding enterprise.

[0067] During the process of uploading the data set to the blockchain, there are multiple preset upload time windows, and each upload time window randomly selects an encryption algorithm for blockchain encryption. The necessity of dividing multiple upload time windows is explained as follows:

[0068] Manage and monitor the blockchain process in stages: By dividing multiple upload time windows, the blockchain process can be monitored and managed more precisely. Information such as the data volume, time consumption, and encryption algorithm performance recorded in each time window can be analyzed separately to identify potential performance issues or anomalies in different time periods. This staged processing method can detect early problems in the blockchain process in a timely manner and prevent the entire system from being dragged down by a single anomaly.

[0069] Improve the stability and fault tolerance of the system: If all data is uploaded within a single time window, once a problem occurs, such as network latency or resource bottlenecks, the entire blockchain process may fail or be interrupted. Dividing multiple time windows can disperse the risks. Even if an anomaly or data quality problem occurs in a certain window, it will not affect the blockchain operations in other time windows, thus improving the stability and fault tolerance of the system.

[0070] Facilitate the evaluation and correction of data quality: By performing the operation of uploading data through multiple time windows, the uploading efficiency values of each window can be calculated separately (such as the average value and standard deviation of the data volume and uploading time), so as to evaluate the data quality. If the quality of a certain time window does not meet the standard, it can be corrected and uploaded again separately without affecting the data processing process of other time windows. This not only improves the accuracy of data management, but also makes the system more flexible in dealing with data anomalies.

[0071] The necessity of randomly obtaining encryption algorithms is explained as follows:

[0072] Improve the security of the system: Randomly obtaining encryption algorithms can increase the difficulty for attackers to crack the system. Because it is difficult for attackers to predict which encryption algorithm will be used in which time window, thus increasing the complexity and cost of attacking the system. By changing the encryption algorithms irregularly, even if a vulnerability of a certain algorithm is discovered by attackers in a specific situation, other randomly selected encryption algorithms will reduce the risk of the overall system being broken.

[0073] Prevent the weaknesses of a single algorithm from being exploited: Each encryption algorithm may have different advantages and weaknesses. On the blockchain, long-term use of the same algorithm may lead to the weaknesses of this algorithm being repeatedly studied and exploited by attackers, thus reducing the security of the system. By randomly obtaining different encryption algorithms, the system vulnerabilities caused by the security defects of a single algorithm can be avoided to a certain extent.

[0074] Disperse the burden of algorithm performance and resources: Different encryption algorithms have different requirements for system resources. Some algorithms may be more intensive in terms of computing performance and memory consumption, while other algorithms are relatively lightweight. By randomly selecting encryption algorithms, the occupation of system resources by algorithms can be balanced in different time windows, optimizing the overall resource allocation and avoiding resource bottlenecks caused by a single algorithm during the data uploading process.

[0075] Performing the analysis of the data uploading time window means:

[0076] Obtain the data volume of each piece of content uploaded within the upload time window, the timestamp at the start, and the timestamp at the end. Subtract the start timestamp from the end timestamp to get the upload time. Then calculate the upload efficiency value for each piece of uploaded content. The upload efficiency value is obtained by dividing the data volume of the uploaded content by the upload time. The upload efficiency value can reflect the time and resource consumption required for each piece of content during the upload process. Next, calculate the average value PJi and the standard deviation BJi of all upload efficiency values within the upload time window; i is the number of the upload time window. The average value can show the overall efficiency level of the upload operation within this window, and the standard deviation reflects the fluctuation of the upload efficiency values. The larger the standard deviation, the less stable the upload efficiency. Each upload time window will be assigned a number to distinguish the upload data situation of different time windows in subsequent analysis.

[0077] When judging whether there are early signs of insufficient quality of uploaded data during the upload in the upload time window according to the analysis result of the upload time window, it refers to:

[0078] Obtain the average value PJi and the standard deviation BJi of all upload efficiency values within the upload time window, and the encryption algorithm randomly obtained for the upload time window i. According to the type of the encryption algorithm, find the corresponding preset standard upload interval 1 and standard upload interval 2. If it is satisfied that the average value PJi falls within the range of standard upload interval 1 and the standard deviation BJi falls within the range of standard upload interval 2, then generate a normal signal. If it is not satisfied that the average value PJi falls within the range of standard upload interval 1 and the standard deviation BJi falls within the range of standard upload interval 2, then generate an abnormal signal. When the abnormal signal is generated, it indicates that there are early signs of insufficient quality of uploaded data in the upload time window. Each upload time window will randomly select an encryption algorithm. According to the selected encryption algorithm, find the corresponding standard upload intervals for this algorithm, which are divided into standard interval 1 and standard interval 2. The two standard intervals are used to evaluate the performance of the upload efficiency values to ensure that the upload process meets the preset encryption security and performance standards. If the average value of the upload efficiency is within the range of standard interval 1 and the standard deviation is within the range of standard interval 2, it means that the upload process is normal and a normal signal is generated; on the contrary, if the average value and the standard deviation are not within the standard interval range, an abnormal signal is generated. The generation of the abnormal signal indicates that there are early signs of insufficient quality of uploaded data in the upload time window. In this way, potential problems in the upload process can be discovered in advance so as to take measures to improve the stability and reliability of data upload.

[0079] The content of feature extraction includes the efficiency information of the encryption algorithm within the upload time window and the adversarial information of the encryption algorithm within the upload time window. Conducting efficiency analysis on the adopted encryption algorithms respectively refers to:

[0080] Obtain the total execution time of the encryption operation, the total number of bytes of the encrypted data, the preset security score corresponding to the encryption algorithm type, the data block size, the average occupancy rate of system resources during the encryption process, and whether the encryption algorithm type supports parallel computing from the efficiency information;

[0081] Then perform the following calculations:

[0082] Tenc represents the total execution time of the encryption operation, that is, the time consumed during the encryption process. The shorter it is, the higher the encryption efficiency. Ddate represents the total number of bytes of the encrypted data, that is, the amount of data to be encrypted. The larger it is, the more likely the encryption time will increase. Ssec represents the preset security score corresponding to the encryption algorithm type. The higher the score, the higher the security of the encryption algorithm. Renc represents the average occupancy rate of computer system resources such as the CPU during the encryption process. The higher the occupancy rate, the more resources are consumed during the encryption process. Penc represents the parallelization efficiency. The higher the parallelization efficiency, the faster the encryption speed may be. When the encryption algorithm type supports parallel computing, Penc takes the value of 1. When the encryption algorithm type does not support parallel computing, Penc takes the value of 0. Bsize is the standard data block size, used to compare the efficiency of the encryption algorithm in processing data blocks, usually a fixed value. Bj is the data block size corresponding to the encryption algorithm type j. Different algorithms may have different data block sizes, which affect the efficiency of the encryption process. For example, the encryption algorithm AES commonly uses 128 bits (16 bytes) as a data block. Aj is the preset adjustment coefficient corresponding to the encryption algorithm type j, used to balance the performance differences of different encryption algorithms. The adjustment coefficients of different algorithms can adjust the calculation result of the encryption efficiency index. Estd is the standard encryption efficiency value, used for comparison and reference. There are a total of N types of encryption algorithms, and the sum of the preset adjustment coefficients corresponding to the N types of encryption algorithms is 1. In a computer system, data is usually stored in binary form. Therefore, many calculations are closely related to binary. The logarithm with base 2 can directly reflect the growth of data blocks in the binary system, which is more in line with the internal operation mechanism of the computer. XLI represents the encryption efficiency index. The larger the encryption efficiency index, the higher the efficiency of the encryption algorithm under the current conditions, that is, it can process a large amount of data in a short time and occupy relatively less resources. An efficient encryption algorithm can optimize resource consumption and execution time while ensuring security, thereby improving the performance and security of the overall system.

[0083] Performing adversarial analysis on the adopted encryption algorithm means:

[0084] Obtain the encryption strength data corresponding to each preset encryption strength index from the adversarial information, the number of network attacks and self - protection times within the on - chain time window, as well as the intensity of each network attack and self - protection; Attack types: brute - force cracking, man - in - the - middle attack, side - channel attack, DDoS attack, etc.; Protection types: key update, firewall enhancement, algorithm upgrade, etc.;

[0085] Then perform the following calculations:

[0086] C is a constant used to adjust the overall encryption confrontation index, while ensuring the denominator part is meaningful and the calculation result is within the expected range. m is the number of the encryption strength index, fm is the encryption strength data corresponding to the encryption strength index m, such as key length, etc. M is the total number of encryption strength indexes, that is, the total number of all different indexes used to measure encryption strength. p is the fluctuation number of the on - chain time window. Fluctuation refers to the security fluctuation caused by network attacks or self - protection. The total number of network attack times and self - protection times is P. kp represents the fluctuation intensity corresponding to the fluctuation number p of the on - chain time window. Fluctuation intensity refers to the network attack intensity or self - protection intensity, representing the intensity of a specific attack or protection event, such as attack intensity or protection effect. vp represents the fluctuation coefficient corresponding to the fluctuation number p of the on - chain time window, reflecting the impact degree of this fluctuation event on the system. The fluctuation coefficient is composed of the negative fluctuation ratio coefficient corresponding to network attacks and the positive fluctuation ratio coefficient corresponding to self - protection, and satisfies that the sum of all negative fluctuation ratio coefficients is - 1, and the sum of all positive fluctuation ratio coefficients is 1. The negative fluctuation ratio coefficient corresponds to the impact of network attacks and is represented by a negative number, and the sum of all negative fluctuation ratio coefficients is - 1. The positive fluctuation ratio coefficient corresponds to the effect of self - protection and is represented by a positive number, and the sum of all positive fluctuation ratio coefficients is 1. CRI represents the encryption confrontation index.

[0087] Represents the basic strength of the encryption system, covering the contributions of different encryption strength indexes. This part reflects the defense ability of the system itself and is the basis of the encryption confrontation index. Through different strength data and weights, the formula can quantify the security performance differences of different encryption algorithms. Used to evaluate the performance of the system under actual attack or protection events. The numerator part combines the intensity and coefficient of all fluctuation events, reflecting the overall impact of each fluctuation event on the system. The denominator part ensures the stability of the calculation result and avoids extreme situations when the numerator is too large. The encryption confrontation index reflects the comprehensive defense ability of the system in the face of external attacks and self - protection behaviors. The larger the encryption confrontation index, the stronger the confrontation ability of the encryption system, that is, under the same conditions, the system can more effectively resist network attacks and maintain its security and stability.

[0088] The logic of fuzzy logic reasoning is as follows: Both the encryption resistance index and the encryption efficiency index of the on-chain time window are used as input variables, and the classification type of the on-chain time window is used as the output variable. The input variables are fuzzified, and the values of the input variables are converted into fuzzy sets. The output variable is fuzzified, and the output variable is converted into a fuzzy set. Fuzzy rules are formulated to describe the on-chain quality under different combinations of data types. The fuzzified input variables are inferred through the fuzzy rules to obtain the classification type of the on-chain time window.

[0089] The encryption resistance index and the encryption efficiency index of the on-chain time window are used as input variables. These variables are respectively fuzzified into two fuzzy sets, namely "high" and "low". For example, the encryption resistance index can be divided into "high resistance" and "low resistance"; the encryption efficiency index can be divided into "high efficiency" and "low efficiency".

[0090] The output variable is the classification type of the on-chain time window. The output variable only contains two fuzzy sets, namely "high-quality window" and "low-quality window", which are used to indicate the quality of the on-chain data. Fuzzy rules applicable to two input variables and two output results are formulated. The specific rules are as follows: If the encryption resistance index is "high" and the encryption efficiency index is "high", it is determined as a "high-quality window". If the encryption resistance index is "low" or the encryption efficiency index is "low", it is determined as a "low-quality window". These rules ensure that the system can automatically evaluate the quality level of the on-chain time window under different combination conditions.

[0091] The fuzzified encryption resistance index and encryption efficiency index are subjected to fuzzy reasoning according to the above rules. For example: When both input variables belong to the "high" set, the system will output the result of "high-quality window". When any one of the input variables is "low", the system will output the result of "low-quality window". Through this logic, simplified fuzzy reasoning can be achieved, and complex encryption conditions and efficiency conditions can be intuitively mapped to the judgment of high- and low-quality windows. The fuzzified result is defuzzified to determine the final classification type, and a clear on-chain time window classification result, namely "high-quality window" or "low-quality window", is output. In the case of this two-window classification, fuzzy logic reasoning mainly judges whether the "high-quality" or "low-quality" conditions are met through preset rules. The system directly determines the quality of the window according to the high and low states of the input variables, making the fuzzy reasoning process more concise and meeting the basic judgment requirements.

[0092] Embodiment 2: A blockchain-based enterprise green credit verification system, including:

[0093] An enterprise data collection module, which collects the respective green behavior data of enterprises from the data sources of multiple enterprises to obtain multiple green behavior data sets, and then aggregates them to obtain an initial data set;

[0094] The verification module performs a preset verification operation based on the initial data collection to ensure the authenticity and integrity of the data, and obtains the verified data collection for uploading to the chain;

[0095] The on-chain analysis module performs an on-chain operation on the data collection for uploading to the chain, and obtains information during the on-chain operation process to get the on-chain process information set and perform an on-chain time window analysis. According to the on-chain time window analysis result, it judges whether there are early signs of insufficient quality of the data uploaded to the chain when uploading within the on-chain time window;

[0096] When there are early signs of insufficient quality of the data uploaded to the chain, the encryption algorithm analysis module extracts features from the on-chain time window, then respectively performs efficiency analysis and adversarial analysis on the encryption algorithms used, and performs fuzzy logic reasoning on the results of the efficiency analysis and adversarial analysis together to divide the on-chain time window into a high-quality on-chain window or a low-quality on-chain window;

[0097] The optimization module performs a re-on-chain operation on the content uploaded to the chain in the data collection for uploading to the chain corresponding to the low-quality on-chain window;

[0098] When conducting a green credit verification of an enterprise, the credit verification module obtains all the green behavior data of the corresponding enterprise from the data uploaded to the chain in all high-quality on-chain windows, then uses a preset green credit scoring model to evaluate the green credit of the enterprise, generates the corresponding green credit score, and sends a green credit warning signal to the corresponding enterprise when the warning mechanism is triggered.

[0099] The specific training steps of the green credit scoring model in the credit verification module are as follows:

[0100] Collect a large number of green behavior data samples from historical records, including indicators such as carbon emissions, energy consumption, waste treatment, and environmental protection certifications. These data come from the high-quality on-chain windows of multiple enterprises to ensure the accuracy and consistency of the data. Assign real green credit scoring labels to each data sample for supervised learning. These scores can be generated according to expert scores or existing standard scores.

[0101] Data preprocessing: Standardize and normalize the original data to eliminate the dimensional differences of different data and improve the stability of model training. Convert the data into a multi-dimensional format suitable for the input of a convolutional neural network, for example, convert time series data into matrix or image form for the convenience of processing by the convolutional layer.

[0102] Construct a convolutional neural network model: Build a convolutional neural network architecture, including convolutional layers, pooling layers, fully connected layers, etc. The convolutional layer is used to extract features of green behavior data, the pooling layer is used for dimensionality reduction to reduce parameters, and the final fully connected layer is used to output the credit score. Design a loss function (such as mean squared error loss) and an optimization algorithm (such as stochastic gradient descent) to minimize the scoring error.

[0103] Model training: Input the preprocessed data into the model, and use the set loss function and optimization algorithm to train the model. Continuously adjust the network parameters through backpropagation, so that the score predicted by the model gradually approaches the true score.

[0104] During the training process, use cross-validation and early stopping techniques to prevent overfitting and ensure the generalization ability of the model.

[0105] Model evaluation: After training is completed, use the test data set to evaluate the prediction performance of the model, and calculate indicators such as the accuracy and mean squared error of the model to ensure that the model can accurately predict the green credit score on unseen data.

[0106] The specific usage steps are as follows: Data input: When conducting the green credit verification of an enterprise, extract the latest green behavior data of the target enterprise from the on-chain data of all high-quality on-chain windows. This data is used as the input of the model to obtain the current green credit score of the enterprise.

[0107] Data preprocessing: Preprocess the new green behavior data to make it conform to the format of the model input, including standardization and normalization processing.

[0108] Model prediction: Input the preprocessed green behavior data into the trained convolutional neural network model. The model evaluates the green behavior of the enterprise through the trained feature extraction and scoring prediction capabilities, and generates a green credit score.

[0109] Score output: Use the green credit score generated by the model as the output for the credit verification module. The score result can be used for the green credit rating of the enterprise. If the score is lower than the set standard, an alarm can also be triggered to remind the enterprise to improve its green behavior.

[0110] The above formulas are all dimensionless and take their numerical values for calculation. The formulas are obtained by collecting a large amount of data for software simulation to get a formula that is closest to the actual situation. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0111] It should be understood that in various embodiments of the present application, the magnitudes of the sequence numbers of the above processes do not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present application.

[0112] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. A professional technician can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of this application.

[0113] Those skilled in the art can clearly understand that for the convenience and conciseness of description, the specific working processes of the systems, devices, and units described above can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0114] As described above, the above is only the specific implementation manner of this application, but the protection scope of this application is not limited thereto. Any person skilled in the art within the technical scope disclosed by this application can easily think of changes or substitutions, which should all be covered by the protection scope of this application. Therefore, the protection scope of this application should be subject to the protection scope of the claims.

Claims

1. A method for enterprise green credit verification based on blockchain, characterized in that It includes the following steps: Collect the respective green behavior data of enterprises from the data sources of multiple enterprises to obtain multiple green behavior data sets, and then summarize them to obtain an initial data set; Perform a preset verification operation based on the initial data set to ensure the authenticity and integrity of the data, and obtain a verified data set for uploading to the chain; Perform an operation of uploading the data set to the chain, and obtain information during the operation of uploading to the chain to obtain an information set of the uploading process and perform an analysis of the uploading time window. According to the result of the analysis of the uploading time window, determine whether there are early signs of insufficient quality of the data uploaded during the uploading time window; When there are early signs of insufficient quality of the data uploaded to the chain, extract the characteristics of the uploading time window, then perform efficiency analysis and adversarial analysis on the encryption algorithms used respectively, perform fuzzy logic reasoning on the results of the efficiency analysis and adversarial analysis together, and divide the uploading time window into a high-quality uploading window or a low-quality uploading window; Perform a re-upload operation on the content uploaded in the low-quality uploading window corresponding to the data set uploaded to the chain; When conducting a green credit verification of an enterprise, obtain all the green behavior data of the corresponding enterprise from the data uploaded in all high-quality uploading windows, and then use a preset green credit scoring model to evaluate the green credit of the enterprise and generate a corresponding green credit score.

2. The enterprise green credit verification method based on blockchain according to claim 1, characterized in that The verification operation includes data source confirmation, data timeliness verification, and data consistency check.

3. The method for verifying the green credit of enterprises based on blockchain according to claim 2, wherein When the green credit score is within the preset standard score range, the early warning mechanism is not triggered. When the green credit score is not within the preset standard score range, the early warning mechanism is triggered, and a green credit warning signal is sent to the corresponding enterprise.

4. The method for verifying the green credit of enterprises based on blockchain according to claim 3, wherein, During the process of performing the operation of uploading the data set to the chain, there are multiple preset uploading time windows, and a random encryption algorithm is obtained for each uploading time window for uploading and encryption.

5. The method for verifying the green credit of enterprises based on blockchain according to claim 4, wherein, Performing an analysis of the uploading time window means: Obtain the data volume of each uploaded content within the uploading time window, the start timestamp, and the end timestamp. Subtract the start timestamp from the end timestamp to obtain the uploading time, and then calculate the uploading efficiency value of each uploaded content. The uploading efficiency value is obtained by dividing the data volume of the uploaded content by the uploading time. Then calculate the average value PJi and the standard deviation BJi of all the uploading efficiency values within the uploading time window; i is the number of the uploading time window.

6. The method for verifying the green credit of an enterprise based on blockchain according to claim 5, wherein Determining whether there are early signs of insufficient quality of the data uploaded during the uploading time window according to the result of the analysis of the uploading time window means: Obtain the average value \(PJi\) and standard deviation \(BJi\) of all the on-chain efficiency values within the on-chain time window, as well as the encryption algorithm randomly obtained within the on-chain time window \(i\). According to the type of the encryption algorithm, find the corresponding preset standard on-chain interval 1 and standard on-chain interval 2. If the average value \(PJi\) falls within the range of standard on-chain interval 1 and the standard deviation \(BJi\) falls within the range of standard on-chain interval 2, then generate a normal signal. If the average value \(PJi\) does not fall within the range of standard on-chain interval 1 and the standard deviation \(BJi\) does not fall within the range of standard on-chain interval 2, then generate an abnormal signal. When the abnormal signal is generated, it indicates an early sign of insufficient on-chain data quality in the on-chain time window.

7. The method for verifying the green credit of enterprises based on blockchain according to claim 6, wherein The content of feature extraction includes the efficiency information of the encryption algorithm within the on-chain time window and the adversarial information of the encryption algorithm within the on-chain time window; Conducting efficiency analysis on the adopted encryption algorithms respectively refers to: Obtain the total execution time of the encryption operation, the total number of bytes of the encrypted data, the preset security score and data block size corresponding to the type of the encryption algorithm, the average occupancy rate of system resources during the encryption process, and whether the type of the encryption algorithm supports parallel computing from the efficiency information; Then perform the following calculations: Tenc represents the total execution time of the encryption operation, Ddate represents the total number of bytes of the encrypted data, Ssec represents the preset security score corresponding to the encryption algorithm type, Renc represents the average occupancy rate of system resources during the encryption process, Penc represents the parallelization efficiency. When the encryption algorithm type supports parallel computing, the value of Penc is 1. When the encryption algorithm type does not support parallel computing, the value of Penc is 0. Bsize is the standard data block size, Bj is the data block size corresponding to the encryption algorithm type j, Aj is the preset adjustment coefficient corresponding to the encryption algorithm type j, Estd is the standard encryption efficiency value for comparison reference. There are a total of N types of encryption algorithms, and the sum of the preset adjustment coefficients corresponding to the N types of encryption algorithms is one. XLI represents the encryption efficiency index.

8. The method for verifying the green credit of an enterprise based on blockchain according to claim 7, wherein, Conducting adversarial analysis on the adopted encryption algorithms respectively refers to: Obtain the encryption intensity data corresponding to each preset encryption intensity index, the number of network attacks and self-protection times within the on-chain time window, and the intensity of each network attack and self-protection intensity from the adversarial information; Then perform the following calculations: C is a constant, m is the serial number of the encryption strength index, fm is the encryption strength data corresponding to the encryption strength index m, M is the total number of encryption strength indexes, p is the fluctuation serial number of the on-chain time window, and the fluctuation refers to the security fluctuation caused by network attacks or self-protection. The total number of network attack times and self-protection times is P, kp represents the fluctuation strength corresponding to the fluctuation serial number p of the on-chain time window, and the fluctuation strength refers to the network attack strength or self-protection strength. vp represents the fluctuation coefficient corresponding to the fluctuation serial number p of the on-chain time window. The fluctuation coefficient is composed of the negative fluctuation ratio coefficient corresponding to the network attack and the positive fluctuation ratio coefficient corresponding to the self-protection, and it satisfies that the sum of all negative fluctuation ratio coefficients is -1 and the sum of all positive fluctuation ratio coefficients is 1. CRI represents the encryption confrontation index.

9. The method for enterprise green credit verification based on blockchain according to claim 8, characterized in that, The logic of fuzzy logic reasoning is: Take both the encryption adversarial index and the encryption efficiency index of the on-chain time window as input variables, and take the division type of the on-chain time window as the output variable. Perform fuzzy processing on the input variables, convert the values of the input variables into fuzzy sets, perform fuzzy processing on the output variable, convert the output variable into a fuzzy set, formulate fuzzy rules to describe the on-chain quality under different combinations of data types, and perform reasoning on the fuzzy input variables through the fuzzy rules to obtain the division type of the on-chain time window.

10. The enterprise green credit verification system based on blockchain is implemented based on the enterprise green credit verification method based on blockchain according to any one of claims 1-9, and is characterized in that, Including: An enterprise data collection module, which collects the respective green behavior data of enterprises from the data sources of multiple enterprises to obtain multiple green behavior data sets, and then summarizes them to obtain an initial data collection; A verification module, which performs preset verification operations based on the initial data collection to ensure the authenticity and integrity of the data, and obtains the verified on-chain data collection; An on-chain analysis module, which performs on-chain operations on the on-chain data collection, and obtains information during the on-chain operation process to obtain an on-chain process information set and perform on-chain time window analysis. According to the on-chain time window analysis result, judge whether there is an early sign of insufficient on-chain data quality when on-chain in the on-chain time window; An encryption algorithm analysis module, when there is an early sign of insufficient on-chain data quality, performs feature extraction on the on-chain time window, then conducts efficiency analysis and adversarial analysis on the adopted encryption algorithms respectively, performs fuzzy logic reasoning on the results of the efficiency analysis and adversarial analysis together, and divides the on-chain time window into a high-quality on-chain window or a low-quality on-chain window; Optimization module, which performs re-upload operations on the uploaded content in the uploaded data set corresponding to the low-quality upload window; Credit verification module, when conducting enterprise green credit verification, obtains all the green behavior data of the corresponding enterprise from the uploaded data of all high-quality upload windows, then uses a preset green credit scoring model to evaluate the green credit of the enterprise, generates the corresponding green credit score, and sends a green credit warning signal to the corresponding enterprise when the warning mechanism is triggered.

Images