Bank fund delivery method and system based on block chain

Abstract

The invention provides a bank fund delivery method and system based on a block chain, and relates to the technical field of finance, and the method comprises the steps: receiving a fund outflow request of a block chain account, and verifying the fund outflow request through a pre-deployed smart contract; after the verification is passed, inputting the transaction characteristics of the associated fund outflow request into a pre-constructed risk assessment model to obtain a risk score; dynamically adjusting a default risk level section of the block chain account based on the transaction characteristics of the fund outflow request to obtain a current risk level section; performing risk grading on the block chain account in combination with the current risk level section and the risk score to obtain a risk level of the block chain account; and searching a pre-constructed level mapping table based on the risk level, determining a fund operation authority level corresponding to the fund outflow request of the block chain account, and executing the fund operation authority level. According to the fund delivery method and system, intelligent, dynamic and secure fund delivery is realized through intelligent contract verification, dynamic risk assessment and level mapping table management.

Description

Technical Field

[0001] The present invention relates to the field of financial technology, and in particular to a bank funds delivery method and system based on blockchain. Background Art

[0002] With the continuous development of blockchain technology, its application in the financial field is becoming more and more extensive, especially in bank fund settlement. Blockchain technology, with its characteristics of decentralization, high transparency, and data immutability, provides a new solution for the safe settlement of funds. However, traditional blockchain fund settlement methods often lack comprehensive assessment and dynamic management of transaction risks, resulting in potential security risks in the fund settlement process.

[0003] In the existing technology, the fund settlement process usually relies on simple transaction verification, such as verifying the basic information such as the account balance and transaction amount of both parties to the transaction. However, this verification method often cannot effectively identify potential risk points, such as the credit status of the counterparty, abnormal transaction frequency, etc. In addition, the traditional fund settlement method lacks a dynamic risk management mechanism and cannot adjust the risk level and fund operation authority level in real time according to transaction characteristics, thus limiting the flexibility and security of fund settlement.

[0004] Therefore, it is necessary to provide a blockchain-based bank funds settlement method and system to solve the above technical problems. Summary of the invention

[0005] In order to solve the above technical problems, the present invention provides a bank funds settlement method and system based on blockchain, which verifies the fund outflow request through a pre-deployed smart contract to ensure the compliance of the transaction; at the same time, a risk assessment model constructed by a machine learning model is used to accurately assess the transaction risk, and the risk level segment of the blockchain account is dynamically adjusted according to the assessment results.

[0006] In addition, the present invention also determines the fund operation authority level according to the risk level through a pre-constructed level mapping table, thereby realizing the intelligent, dynamic and secure fund delivery.

[0007] The present invention provides a bank funds delivery method based on blockchain, the method comprising the following steps:

[0008] Receive a fund outflow request from a blockchain account and verify the fund outflow request using a pre-deployed smart contract;

[0009] After verification, the transaction features associated with the fund outflow request are input into a pre-built risk assessment model to obtain a risk score, wherein the risk assessment model is a machine learning model trained based on historical transaction data;

[0010] Dynamically adjust the default risk level segment of the blockchain account based on the transaction characteristics of the fund outflow request to obtain a current risk level segment;

[0011] Performing risk grading on the blockchain account in combination with the current risk level segment and the risk score to obtain a risk level of the blockchain account;

[0012] Based on the risk level, a pre-built level mapping table is searched to determine and execute the fund operation authority level corresponding to the fund outflow request of the blockchain account.

[0013] Preferably, the receiving a fund outflow request from a blockchain account and verifying the fund outflow request using a pre-deployed smart contract includes:

[0014] Determining whether the fund outflow request meets the preset transaction rules through a smart contract, wherein the preset transaction rules include transaction amount limit, counterparty whitelist, transaction time window limit and transaction frequency limit;

[0015] Confirm through a smart contract whether the number of nodes recognized by the fund outflow request in the blockchain network reaches a threshold, wherein the threshold is dynamically determined based on the transaction characteristics of the fund outflow request.

[0016] Preferably, judging whether the fund outflow request satisfies a preset transaction rule through a smart contract includes:

[0017] a. Determine whether the transaction amount of the fund outflow request exceeds the set maximum transaction limit;

[0018] If the transaction amount exceeds the maximum transaction limit, reject the fund outflow request;

[0019] b. Check whether the recipient address of the fund outflow request is in the preset whitelist of the account;

[0020] If the recipient address is not in the whitelist, the fund outflow request is rejected;

[0021] c. Confirm whether the time point of the fund outflow request is within the set permitted trading time period;

[0022] If the transaction time point is not within the permitted time period, the fund outflow request is rejected;

[0023] d. Count the number of fund outflows from the account within the set time window;

[0024] Determine whether the number of fund outflows from the blockchain account exceeds a set maximum frequency;

[0025] If the number of fund outflows from the blockchain account exceeds the maximum frequency, the fund outflow request is rejected.

[0026] Preferably, the step of confirming through a smart contract whether the number of nodes in the blockchain network that have recognized the fund outflow request has reached a threshold includes:

[0027] Dynamically determine a threshold value based on the transaction characteristics of the fund outflow request;

[0028] Confirming whether the number of nodes in the blockchain network that have recognized the fund outflow request has reached the dynamically determined threshold;

[0029] If the number of nodes does not exceed the threshold, rejecting the fund outflow request;

[0030] If the number of nodes exceeds the threshold, proceed to the next step.

[0031] Preferably, the method for dynamically determining the threshold value includes:

[0032] Assign a weight to each transaction feature, wherein the transaction features include transaction amount A, user credit score C, transaction frequency F, transaction time T, and transaction location L;

[0033] The comprehensive risk score of the fund outflow request is calculated based on the weight, wherein the calculation formula of the comprehensive risk score is:

[0034] R=ω A ·A+ω C ·C+ω F ·F+ω T ·T+ω L ·L

[0035] Among them, R represents the comprehensive risk score, ω A ,ω C ,ω F ,ω T and ω L They represent the weights of transaction amount A, user credit score C, transaction frequency F, transaction time T and transaction location L respectively, and their sum is 1;

[0036] The threshold is dynamically determined according to the comprehensive risk score, wherein the formula for determination is:

[0037] N=N0+k·R

[0038] Wherein N0 represents a preset basic threshold, k represents an adjustment coefficient, and N represents a determined threshold.

[0039] Preferably, the process of constructing the risk assessment model includes:

[0040] Collect historical transaction data from the blockchain network;

[0041] Using historical transaction data to train a risk assessment model to minimize prediction errors, wherein the risk assessment model uses an XGBoost algorithm;

[0042] Deploy the trained risk assessment model based on the XGBoost algorithm to the blockchain network.

[0043] Preferably, dynamically adjusting the default risk level segment of the blockchain account based on the transaction characteristics of the fund outflow request to obtain the current risk level segment includes:

[0044] According to the attribute of the blockchain account, searching for a default risk level segment of the blockchain account from a pre-built risk level segment mapping table, wherein the risk level segment mapping table includes default risk level segments corresponding to different attributes;

[0045] According to the comprehensive risk score, the intervals in the default risk level segment are adjusted in sections to obtain the current risk level segment.

[0046] Preferably, the risk grading is performed based on a risk grading rule, wherein the risk grading rule comprises a plurality of risk levels and corresponding scoring ranges, and the risk levels include low risk, medium risk and high risk.

[0047] Preferably, searching a pre-built level mapping table based on the risk level, determining and executing the fund operation permission level corresponding to the fund outflow request of the blockchain account, includes:

[0048] According to the risk level of the blockchain account, searching from a pre-built level mapping table to obtain the fund operation permission level corresponding to the risk level, wherein the level mapping table contains fund operation permission levels corresponding to different risk levels;

[0049] According to the determined fund operation authority level, the operation corresponding to the fund operation authority level is performed.

[0050] The present invention also provides a blockchain-based bank funds settlement system for executing the blockchain-based bank funds settlement method, the system comprising:

[0051] A transaction verification module, used to receive a fund outflow request from a blockchain account and verify the fund outflow request using a pre-deployed smart contract;

[0052] A risk assessment module, configured to input the transaction features associated with the fund outflow request into a pre-built risk assessment model to obtain a risk score after verification, wherein the risk assessment model is a machine learning model trained based on historical transaction data;

[0053] A risk level adjustment module, configured to dynamically adjust a default risk level segment of the blockchain account based on the transaction characteristics of the fund outflow request to obtain a current risk level segment;

[0054] A risk grading module, used to perform risk grading on the blockchain account in combination with the current risk level segment and the risk score to obtain the risk level of the blockchain account;

[0055] The level verification module is used to search a pre-built level mapping table based on the risk level, determine the fund operation authority level corresponding to the fund outflow request of the blockchain account, and execute it.

[0056] Compared with the related art, the bank funds settlement method and system based on blockchain provided by the present invention have the following beneficial effects:

[0057] The present invention uses smart contracts to determine whether the fund outflow request meets the preset transaction rules, including transaction amount limit, counterparty whitelist, transaction time window limit and transaction frequency limit, so as to ensure the compliance and security of the transaction.

[0058] At the same time, the present invention uses a risk assessment model constructed by a machine learning model to accurately assess transaction risks based on historical transaction data, and dynamically adjusts the risk level segment of the blockchain account based on the assessment results. This dynamic risk assessment mechanism can reflect changes in transaction risks in real time and improve the security of fund delivery.

[0059] In addition, through the pre-built level mapping table, the fund operation permission level is determined according to the risk level of the blockchain account, thus realizing the intelligent management of fund delivery. This management method can adopt corresponding fund use strategies according to different risk levels, and improve the flexibility and efficiency of fund delivery.

[0060] In summary, the present invention proposes a bank funds settlement method based on blockchain, which realizes the intelligent, dynamic and secure funds settlement through technical means such as smart contract verification, dynamic risk assessment and level mapping table management. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 A flowchart of a bank funds delivery method based on blockchain provided by the present invention;

[0062] Figure 2A schematic diagram of the module structure of a blockchain-based bank funds settlement system provided by the present invention. DETAILED DESCRIPTION

[0063] The present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are only used to explain the present invention, rather than to limit the present invention. It should also be noted that, for ease of description, only the parts related to the present invention, rather than all structures, are shown in the accompanying drawings. In addition, the embodiments of the present invention and the features in the embodiments may be combined with each other without conflict.

[0064] It should also be noted that, for ease of description, only the part relevant to the present invention but not all content is shown in the accompanying drawings. It should be mentioned before discussing exemplary embodiments in more detail that some exemplary embodiments are described as processing or methods depicted as flow charts. Although the flow chart describes each operation (or step) as sequential processing, many operations therein can be implemented in parallel, concurrently or simultaneously. In addition, the order of each operation can be rearranged. When its operation is completed, the processing can be terminated, but it can also have additional steps not included in the accompanying drawings. The processing can correspond to methods, functions, procedures, subroutines, subprograms, etc.

[0065] Embodiment 1

[0066] It should be noted that the blockchain network of this application is a decentralized network composed of multiple nodes. These nodes jointly maintain a distributed ledger. Each node saves a complete transaction record and ensures the consistency and security of transactions through a consensus mechanism. It has the following functions:

[0067] When a user initiates a fund outflow request through a blockchain account, the request will be broadcast to the entire blockchain network and all nodes will receive the request. The node verifies the fund outflow request through a smart contract to ensure that the transaction complies with the preset rules. If the transaction does not comply with the rules, the smart contract will automatically reject the request.

[0068] The nodes reach a consensus through the consensus mechanism, confirm the validity of the transaction, and record the transaction in the blockchain. The consensus mechanism ensures the immutability and transparency of the transaction.

[0069] Based on the risk level of the blockchain account, the node will execute the corresponding level of fund operation authority to ensure the security and compliance of the transaction.

[0070] All transaction records and permission execution results will be recorded in the blockchain to ensure the transparency and traceability of the system.

[0071] A blockchain account is a user's unique identifier in the blockchain network and is used to store the user's digital assets and transaction records. Each blockchain account has a unique public key and private key pair. Users sign transaction requests with private keys to ensure the security and immutability of transactions. It has the following functions:

[0072] Users can initiate fund outflow requests (such as transfers, payments, etc.) through their blockchain accounts and broadcast the requests to the blockchain network.

[0073] Users can receive funds from other accounts through their blockchain accounts.

[0074] Users can view and manage their digital assets, including balance inquiries, transaction history, etc.

[0075] In summary, the interaction between blockchain accounts and blockchain networks is mainly achieved through smart contracts. After a user initiates a request for fund outflow through a blockchain account, the request will be broadcast to the blockchain network and all nodes will receive the request. The node verifies the fund outflow request through the smart contract to ensure that the transaction complies with the preset rules. If the transaction meets the conditions, the smart contract will approve the request and record the transaction in the blockchain.

[0076] The present invention provides a bank funds delivery method based on blockchain, referring to Figure 1 As shown, the method comprises the following steps:

[0077] S1: Receive a fund outflow request from a blockchain account and verify the fund outflow request using a pre-deployed smart contract.

[0078] Specifically, step S1 includes the following steps:

[0079] S11: Determine whether the fund outflow request meets the preset transaction rules through the smart contract, wherein the preset transaction rules include transaction amount limit, counterparty whitelist, transaction time window limit and transaction frequency limit.

[0080] In the blockchain-based bank fund delivery method, it is crucial to ensure the security and compliance of each fund outflow request. To achieve this goal, the fund outflow request must first be strictly verified. This step is completed through smart contracts, which are automated programs deployed on the blockchain network that can automatically perform specific operations when preset conditions are met. One of the core functions of smart contracts is to verify transaction requests to ensure that transactions comply with preset rules. These rules include but are not limited to transaction amount limits, counterparty whitelists, transaction time window limits, and transaction frequency limits. Through the verification mechanism of smart contracts, illegal or high-risk transactions can be effectively prevented from entering the subsequent processing process, thereby ensuring the security and stability of the system.

[0081] When a user initiates a fund outflow request through a blockchain account, the request will be broadcast to the entire blockchain network and all nodes will receive the request. The smart contract will automatically verify the request. The specific verification process for each node is as follows:

[0082] a. Transaction amount verification: The smart contract first checks whether the transaction amount of the fund outflow request exceeds the set maximum transaction limit. If the transaction amount exceeds the maximum transaction limit, the smart contract will automatically reject the request and return the rejection reason to the user. This mechanism can prevent the occurrence of large abnormal transactions and reduce potential financial risks.

[0083] b. Recipient address verification: The smart contract then checks whether the recipient address of the fund outflow request is in the account's preset whitelist. The whitelist usually contains verified and trusted counterparty addresses. If the recipient address is not in the whitelist, the smart contract will reject the request and notify the user. This mechanism can prevent funds from flowing to unknown or untrusted accounts and enhance the security of transactions.

[0084] c. Transaction time window verification: The smart contract will also confirm whether the time point of the fund outflow request is within the set allowed trading time period. For example, some accounts may only be allowed to trade between 8:00 and 18:00 on weekdays. If the transaction time point is not within the allowed time period, the smart contract will reject the request and inform the user. This mechanism can prevent transactions during abnormal time periods and reduce potential risks.

[0085] d. Transaction frequency verification: The smart contract will count the number of fund outflows from the account within the set time window (such as the past 24 hours) and determine whether it exceeds the set maximum frequency. If the number of fund outflows from the account exceeds the maximum frequency, the smart contract will reject the request and notify the user. This mechanism can prevent frequent abnormal trading behaviors and protect the user's account security.

[0086] In order to ensure that the fund outflow request is recognized by the current node, the following conditions must be met:

[0087] All four verifications passed: Only when all four verification steps above are passed, the node will mark the request as "approved". If any verification step fails, the node will immediately mark the request as "disapproved" and terminate subsequent verification.

[0088] Feedback mechanism: Once a node completes all verifications, it sends a feedback signal to other nodes in the blockchain network, indicating whether the node approves the fund outflow request. The feedback signal contains the following information:

[0089] Node ID: Identifies which node the feedback signal comes from.

[0090] Verification result: indicates whether the node approves the request ("approved" or "disapproved").

[0091] Rejection reason (if any): If the node rejects the request, the feedback signal will include the specific rejection reason (such as the transaction amount exceeds the limit, the recipient address is not in the whitelist, etc.).

[0092] S12: Confirming, through a smart contract, whether the number of nodes in the blockchain network that recognize the fund outflow request has reached a threshold, wherein the threshold is dynamically determined based on the transaction characteristics of the fund outflow request.

[0093] After the outflow request has passed the initial verification of the smart contract, it is necessary to further confirm whether the request has been sufficiently recognized in the blockchain network. This is achieved through the consensus mechanism. The consensus mechanism is one of the core mechanisms in the blockchain network, which ensures that all nodes agree on the validity of the transaction. In order to ensure the security and reliability of the transaction, it is necessary to confirm whether the number of nodes that have recognized the outflow request in the blockchain network has reached a threshold. This threshold is not fixed, but is dynamically determined based on the characteristics of the transaction. In this way, the number of recognized nodes can be adjusted according to different transaction characteristics to ensure that high-risk transactions receive more verification, thereby improving the security and reliability of the system.

[0094] More specifically, the process of confirming through a smart contract whether the number of nodes in the blockchain network that have recognized the fund outflow request has reached a threshold includes:

[0095] Step 1: Dynamically determine a threshold value based on the transaction characteristics of the fund outflow request.

[0096] The method for dynamically determining the threshold value includes the following steps:

[0097] First, a weight is assigned to each transaction feature, wherein the transaction features include transaction amount A, user credit score C, transaction frequency F, transaction time T, and transaction location L.

[0098] In this embodiment, weights are assigned to each feature based on the transaction features of the fund outflow request (transaction amount, user credit score, transaction frequency, transaction time, and transaction location). These weights reflect the degree of impact of different features on transaction risk. The following is a detailed description of each transaction feature and its weight assignment:

[0099] 1. Transaction amount

[0100] The transaction amount is one of the key factors in assessing transaction risk. Generally speaking, transactions with larger transaction amounts are usually considered to be higher risk, as large transactions may involve money laundering, fraud or other illegal activities. Therefore, the system will give a higher weight to the transaction amount.

[0101] According to preset rules or historical data analysis, a reasonable weight range is determined. For example, for ordinary accounts, the weight of the transaction amount may be set to 0.3 to 0.5, depending on the risk level of the account and historical trading behavior. If the historical transaction records of an account show that it often conducts small transactions, then when the account suddenly initiates a large transaction, it will be considered a high-risk transaction and the weight of the transaction amount will be increased accordingly.

[0102] Example: Assume that the default weight distribution of an account is as follows:

[0103] Small transactions (<1,000 yuan): weight is 0.2.

[0104] Medium-amount transactions (1,000 to 5,000 yuan): The weight is 0.4.

[0105] Large transactions (>5,000 yuan): weight is 0.6.

[0106] If a user initiates a transaction of 8,000 yuan, the weight of the transaction amount will be set to 0.6 because it belongs to the large transaction category and has a higher risk.

[0107] 2. User credit score

[0108] User credit score is an important indicator for evaluating user credibility. Users with higher credit scores are generally considered low-risk users because they have a good credit record and stable transaction behavior. Conversely, users with lower credit scores may be considered higher risk, especially when they initiate large or frequent transactions.

[0109] The weight is adjusted dynamically based on the user's credit score. Users with higher credit scores will have a relatively lower risk weight for their transactions; while users with lower credit scores will have their risk weight for their transactions increased accordingly. This mechanism can effectively prevent users with poor credit from making high-risk transactions.

[0110] Example: Assuming an account has a credit score of 750 out of 1000, the weights are assigned according to the following rules:

[0111] Credit score ≥ 800: weight is 0.1.

[0112] Credit score between 600 and 800: weighting is 0.3.

[0113] Credit score < 600: weight is 0.5.

[0114] If the user's credit score is 750, the system will set the weight of the user's credit score to 0.3, indicating that the user's risk is at a medium level.

[0115] 3. Trading frequency

[0116] Transaction frequency refers to the number of fund outflows from the account within a set time window (such as the past 24 hours or 7 days). Frequent trading behavior may indicate that the user is performing abnormal operations, especially when a large number of transactions are initiated in a short period of time, which may pose a risk of money laundering, cashing out or other illegal activities. Therefore, a certain weight will be given to the transaction frequency to assess the potential risk of the transaction.

[0117] The weight is adjusted dynamically based on the account's historical transaction frequency and current transaction frequency. If an account initiates too many transactions in a short period of time, it will be considered a high-risk transaction and the weight of the transaction frequency will be increased accordingly. On the contrary, if the account's transaction frequency remains within the normal range, the weight will be relatively low.

[0118] Example: Assume that an account has initiated 5 transactions in the past 24 hours, and the historical average transaction frequency of the account is 2 transactions per day. The system will assign weights according to the following rules:

[0119] ≤ 2 transactions per day: weight is 0.1.

[0120] 3 to 5 trades per day: weighting is 0.3.

[0121] > 5 trades per day: weight is 0.5.

[0122] In this case, the weight of the transaction frequency is set to 0.3, indicating that the transaction frequency of this account is slightly higher than normal and there is a certain risk.

[0123] 4. Trading hours

[0124] The transaction time refers to the specific time when the fund outflow request occurs. Transactions during certain time periods may be considered high-risk. For example, transactions during non-working hours (such as late at night or on holidays) may not be consistent with normal trading behavior and may be at risk of fraud or other illegal activities. Therefore, a certain weight will be given to the transaction time to assess the potential risk of the transaction.

[0125] The weight is adjusted dynamically based on the time period in which the transaction occurs. For example, transactions between 8:00 and 18:00 on weekdays are generally considered low risk, while transactions late at night (such as 22:00 to 6:00 the next day) or on holidays may be considered high risk. The system will adjust the weight of the transaction time based on these time periods.

[0126] Example: If an account initiates a transaction at 3:00 AM, the weight will be assigned according to the following rules:

[0127] Weekdays 8:00 to 18:00: Weight is 0.1.

[0128] Working days from 18:00 to 22:00 or the next day from 6:00 to 8:00: The weight is 0.3.

[0129] Late night (22:00 to 6:00 the next day) or holidays: the weight is 0.5.

[0130] In this case, the weight of the transaction time is set to 0.5, indicating that the transaction occurred in a high-risk time period and may have potential risks.

[0131] 5. Trading location

[0132] The transaction location refers to the geographical location where the fund outflow request occurs. Transactions in certain regions may be considered higher risk, especially when the transaction location is inconsistent with the user's permanent address, there may be risks of cross-border money laundering, fraud or other illegal activities. Therefore, a certain weight will be given to the transaction location to assess the potential risk of the transaction.

[0133] The weight is dynamically adjusted based on the relationship between the transaction location and the user's permanent address. If the transaction location is consistent with the user's permanent address, the weight will be relatively low; if the transaction location is inconsistent with the user's permanent address, especially when the transaction location is in a high-risk area, the weight will increase accordingly.

[0134] Example: Assume that a user's permanent address is Beijing, China, and the user initiates a transaction from the United States. The weight will be assigned according to the following rules:

[0135] The transaction location is consistent with the permanent address: the weight is 0.1.

[0136] The transaction location is inconsistent with the permanent address, but is located in a low-risk area: the weight is 0.3.

[0137] The transaction location is inconsistent with the permanent address and is located in a high-risk area: the weight is 0.5.

[0138] In this case, the weight of the transaction location is set to 0.5, indicating that the transaction occurred in a high-risk area and there may be potential risks.

[0139] Secondly, the comprehensive risk score of the fund outflow request is calculated based on the weight, wherein the calculation formula of the comprehensive risk score is:

[0140] R=ω A ·A+ω C ·C+ω F ·F+ω T ·T+ω L ·L

[0141] Among them, R represents the comprehensive risk score, ω A ,ω C ,ω F ,ω T and ω L They represent the weights of transaction amount A, user credit score C, transaction frequency F, transaction time T and transaction location L respectively, and the sum is 1.

[0142] In this embodiment, the comprehensive risk score reflects the overall risk level of the transaction.

[0143] Finally, the threshold is dynamically determined according to the comprehensive risk score, wherein the formula for determination is:

[0144] N=N0+k·R

[0145] Wherein N0 represents a preset basic threshold, k represents an adjustment coefficient, and N represents a determined threshold.

[0146] In this embodiment, the basic threshold N0 is a pre-set minimum number of approved nodes, and the adjustment coefficient k is used to dynamically adjust the threshold according to the comprehensive risk score. For high-risk transactions, more nodes will be required to approve the request to ensure the security of the transaction; for low-risk transactions, the number of approved nodes can be appropriately reduced to improve the efficiency of transaction processing.

[0147] Step 2: confirm whether the number of nodes in the blockchain network that have recognized the fund outflow request has reached the dynamically determined threshold;

[0148] If the number of nodes does not exceed the threshold, rejecting the fund outflow request;

[0149] If the number of nodes exceeds the threshold, proceed to the next step.

[0150] In this embodiment, after the smart contract completes the initial verification, all nodes will vote on the fund outflow request according to the consensus mechanism. Each node will send a feedback signal to the blockchain network based on its own verification results, indicating whether the node approves the request. Similarly, the feedback signal contains the following information:

[0151] Node ID: Identifies which node the feedback signal comes from.

[0152] Verification result: indicates whether the node approves the request ("approved" or "disapproved").

[0153] Rejection reason (if any): If the node rejects the request, the feedback signal will include the specific rejection reason (such as the transaction amount exceeds the limit, the recipient address is not in the whitelist, etc.).

[0154] The smart contract will then collect feedback signals from all nodes and count the number of nodes that approve the request. Only when the number of approved nodes exceeds a dynamically determined threshold will the smart contract proceed to the next step. If the number of approved nodes does not reach the threshold, the smart contract will reject the request and notify the user.

[0155] If the number of approved nodes does not exceed the dynamically determined threshold, the smart contract will reject the fund outflow request and return the rejection reason to the user. This mechanism can prevent malicious attacks and tampering, ensuring the security and reliability of transactions.

[0156] If the number of approved nodes exceeds a dynamically determined threshold, the smart contract approves the transaction and records it in the blockchain. At this point, the transaction will enter the subsequent risk assessment and authority management process to ensure the compliance and security of the transaction.

[0157] S2: After the verification is passed, the transaction features associated with the fund outflow request are input into a pre-built risk assessment model to obtain a risk score, wherein the risk assessment model is a machine learning model trained based on historical transaction data.

[0158] In this embodiment, in the blockchain-based bank funds delivery method, it is crucial to ensure the security and compliance of each fund outflow request. Although the smart contract has preliminarily verified the transaction in step S1, in order to further enhance the security of the system, a more in-depth risk assessment of the transaction is required. This step is completed through a pre-built risk assessment model, which can predict the risk level of each transaction based on historical transaction data and give a corresponding risk score. The introduction of the risk assessment model can not only help the system identify high-risk transactions more accurately, but also provide a basis for the subsequent fund operation authority level management to ensure the compliance and security of the transaction.

[0159] In step S2, a risk assessment model is also constructed, and the specific construction process includes the following steps:

[0160] Collect historical transaction data from the blockchain network.

[0161] In this embodiment, in order to build an effective risk assessment model, it is first necessary to collect a large amount of historical transaction data from the blockchain network. These data include but are not limited to transaction amounts, user credit scores, transaction frequencies, transaction times, transaction locations and other features. Historical transaction data is the basis for training the model. By analyzing these data, the model can learn the correlation between different features and transaction risks, thereby improving the accuracy of predictions.

[0162] The risk assessment model is trained using historical transaction data to minimize the prediction error, wherein the risk assessment model adopts the XGBoost algorithm.

[0163] In this embodiment, the XGBoost algorithm is used to train the risk assessment model. XGBoo st (ExtremeGradientBoosting) is a machine learning algorithm based on gradient boosting decision tree (GBDT), which has the ability to process large-scale data and can generate efficient prediction models in a short time. The advantage of XGBoost is that it can automatically handle missing values, prevent overfitting, and can accelerate the training process through parallel computing.

[0164] During training, the collected historical transaction data is used as input to train the XGBoost model. The goal of training is to minimize the prediction error, that is, to make the model predict the risk level of each transaction as accurately as possible. To achieve this goal, the historical transaction data is divided into a training set and a test set. The training set is used to train the model, while the test set is used to evaluate the performance of the model. By continuously adjusting the parameters of the model, the system can gradually optimize the model's prediction ability.

[0165] During the training process, feature engineering is performed on the transaction data to extract features that are useful for risk assessment. For example, the average transaction amount, maximum transaction amount, transaction frequency, transaction time distribution, and other features of each user are calculated. These features will be used as input variables of the model to help the model better understand the risk characteristics of transactions.

[0166] Deploy the trained risk assessment model based on the XGBoost algorithm to the blockchain network.

[0167] In this embodiment, once the risk assessment model is trained and verified by the test set, the model is deployed to the blockchain network. The deployed model can automatically perform risk assessment on transactions after each fund outflow request passes preliminary verification. Specifically, when a user initiates a fund outflow request, the smart contract inputs the characteristics of the transaction into the risk assessment model, and the model outputs a risk score indicating the risk level of the transaction.

[0168] S3: Dynamically adjust the default risk level segment of the blockchain account based on the transaction characteristics of the fund outflow request to obtain a current risk level segment.

[0169] In the blockchain-based bank fund delivery method, it is crucial to ensure the security and compliance of each fund outflow request. To achieve this goal, it is necessary to dynamically adjust the risk level segment of the user based on the user's transaction behavior and account attributes. This step is completed through a pre-built risk level segment mapping table. The risk level segment mapping table is a pre-set rule base that corresponds different account attributes to the default risk level segment. In this way, the initial risk level segment of the account can be quickly determined each time a fund outflow request occurs, providing a basis for subsequent risk assessment.

[0170] Specifically, step S3 includes the following steps:

[0171] S31: According to the attributes of the blockchain account, a default risk level segment of the blockchain account is searched from a pre-constructed risk level segment mapping table, wherein the risk level segment mapping table includes default risk level segments corresponding to different attributes.

[0172] In this embodiment, a risk level segment mapping table is pre-built according to different account attributes (user credit scores). This mapping table contains multiple risk level segments, each segment corresponding to a different risk level. For example, a low-risk segment may be suitable for users with high credit scores and good historical transaction records; while a high-risk segment is suitable for users with low credit scores or abnormal transaction behaviors.

[0173] For example, it is assumed that three risk level segments are preset:

[0174] Low-risk segment: Applicable to users with a credit score ≥ 800 and normal historical transaction behavior.

[0175] Medium-risk segment: Applicable to users with credit scores between 600 and 800 who occasionally make small abnormal transactions.

[0176] High-risk segment: Applicable to users with a credit score <600 and who frequently conduct large or abnormal transactions.

[0177] These segments are matched with the corresponding account attributes in the mapping table so that each time a fund outflow request occurs, the default risk level segment of the account can be quickly looked up and determined.

[0178] When a user initiates a fund outflow request, the smart contract will first look up the default risk level segment of the account from the pre-built risk level segment mapping table based on the user's credit score. Specifically, the smart contract will check the account's credit score and find the corresponding default risk level segment in the mapping table based on the score.

[0179] S32: According to the comprehensive risk score, the intervals in the default risk level segment are adjusted in sections to obtain the current risk level segment.

[0180] In this embodiment, in step S12, a comprehensive risk score has been calculated based on the transaction characteristics of the capital outflow request. This comprehensive risk score reflects the overall risk level of the transaction and is an important basis for dynamically adjusting the risk level segment.

[0181] Suppose a user initiates a transaction of 8,000 yuan. The smart contract calculates a comprehensive risk score of 0.75 based on the transaction characteristics, indicating that the transaction is highly risky.

[0182] Then, based on the comprehensive risk score, the intervals in the default risk level segment are adjusted in sections to obtain the current risk level segment.

[0183] Different adjustment rules are set according to different ranges of comprehensive risk scores. For example, if the comprehensive risk score is between 0.6 and 0.8, the risk level segment will be raised by one level; if the comprehensive risk score exceeds 0.8, the risk level segment will be raised by two levels. This segmented adjustment rule can flexibly adjust the risk level of the account according to the risk level of the transaction, ensuring that high-risk transactions receive more attention.

[0184] Assume that the following segmented adjustment rules are set:

[0185] Comprehensive risk score ≤ 0.4: Keep the default risk level range unchanged.

[0186] 0.4<comprehensive risk score≤0.7: raise the risk level segment by one level.

[0187] Comprehensive risk score>0.7: the risk level segment will be increased by two levels.

[0188] If the comprehensive risk score of an account is 0.75, according to the above rules, its risk level segment will be raised by two levels, from "medium risk segment" to "high risk segment".

[0189] S4: Risk grading the blockchain account in combination with the current risk level segment and the risk score to obtain the risk level of the blockchain account.

[0190] In step S4, the risk grading is performed based on a risk grading rule, wherein the risk grading rule includes multiple risk levels and corresponding scoring ranges, and the risk levels include low risk, medium risk and high risk.

[0191] In this embodiment, in step S3, the default risk level segment of the blockchain account has been dynamically adjusted according to the comprehensive risk score to obtain the current risk level segment. Next, the blockchain account will be risk graded in combination with the current risk level segment and the risk score output by the risk assessment model. Specifically, according to the preset risk grading rules, the risk score of the account is mapped to different risk levels (such as low risk, medium risk, and high risk).

[0192] A set of risk grading rules are predefined to map risk scores to different risk levels. These rules usually contain multiple risk levels and corresponding score ranges. For example:

[0193] Low risk: Risk score < 0.3.

[0194] Medium risk: 0.3≤risk score<0.7.

[0195] High risk: risk score ≥ 0.7.

[0196] Example: If the risk score of an account is 0.8, according to the above rules, the account will be classified as "high risk". If the risk score of another account is 0.4, the account will be classified as "medium risk".

[0197] According to the risk grading rules, the account's risk score is mapped to the corresponding risk level and used as the final risk level of the account. This risk level will be used for subsequent fund operation permission level management to ensure the compliance and security of transactions.

[0198] S5: Searching a pre-built level mapping table based on the risk level, determining and executing the fund operation permission level corresponding to the fund outflow request of the blockchain account.

[0199] Specifically, step S5 includes the following steps:

[0200] S51: According to the risk level of the blockchain account, a fund operation permission level corresponding to the risk level is searched from a pre-built level mapping table, wherein the level mapping table contains fund operation permission levels corresponding to different risk levels.

[0201] In this embodiment, according to the risk level of the blockchain account, the fund operation permission level corresponding to the risk level is searched from the pre-built level mapping table. The level mapping table is a predefined table that contains the fund operation permission levels corresponding to different risk levels (such as low risk, medium risk, and high risk). The system will automatically match the corresponding level according to the risk level of the account.

[0202] The level mapping table is pre-built and usually developed by banks or financial institutions based on regulatory requirements and internal risk control policies.

[0203] According to the risk level of the account, the corresponding level is automatically matched from the level mapping table and applied to the fund outflow request of the account. This automated mechanism not only improves the efficiency of the system, but also reduces the need for manual intervention and reduces operational risks.

[0204] S52: According to the determined fund operation authority level, perform operations corresponding to the fund operation authority level.

[0205] In this embodiment, according to the identity authentication information provided by the blockchain account, its fund operation permission level is retrieved, and the user's fund operation permission level is compared with the permission level required for the fund operation to be performed. If the user's fund operation permission level meets or exceeds the permission level required for the fund operation to be performed, the operation is allowed to continue. If the user's fund operation permission level is insufficient, the system rejects the operation and returns an error message of insufficient permission to the user.

[0206] Embodiment 2

[0207] The present invention also provides a blockchain-based bank funds settlement system for executing the blockchain-based bank funds settlement method. Figure 2 As shown, the system comprises:

[0208] The transaction verification module 100 is used to receive a fund outflow request from a blockchain account and verify the fund outflow request using a pre-deployed smart contract.

[0209] The risk assessment module 200 is used to input the transaction features associated with the fund outflow request into a pre-built risk assessment model to obtain a risk score after verification, wherein the risk assessment model is a machine learning model trained based on historical transaction data.

[0210] The risk level adjustment module 300 is used to dynamically adjust the default risk level segment of the blockchain account based on the transaction characteristics of the fund outflow request to obtain a current risk level segment.

[0211] The risk grading module 400 is used to perform risk grading on the blockchain account in combination with the current risk level segment and the risk score to obtain the risk level of the blockchain account.

[0212] The level verification module 500 is used to search a pre-built level mapping table based on the risk level, determine the fund operation authority level corresponding to the fund outflow request of the blockchain account, and execute it.

[0213] The present application is described with reference to flowcharts and / or block diagrams of methods, devices (systems) and computer program products according to embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0214] Those skilled in the art will appreciate that all or part of the steps in the various methods of the above embodiments may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, the storage medium including a read-only memory (ROM), a random access memory (RAM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), a one-time programmable read-only memory (OTPROM), an electronically erasable programmable read-only memory (EEPROM), a compact disc (CD-ROM) or other optical disc storage, magnetic disk storage, magnetic tape storage, or any other computer-readable medium that can be used to carry or store data.

[0215] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

Claims

1. A bank funds settlement method based on blockchain, characterized in that: The method comprises the following steps: Receive a fund outflow request from a blockchain account and verify the fund outflow request using a pre-deployed smart contract; After verification, the transaction features associated with the fund outflow request are input into a pre-built risk assessment model to obtain a risk score, wherein the risk assessment model is a machine learning model trained based on historical transaction data; Dynamically adjust the default risk level segment of the blockchain account based on the transaction characteristics of the fund outflow request to obtain a current risk level segment; Performing risk grading on the blockchain account in combination with the current risk level segment and the risk score to obtain a risk level of the blockchain account; Based on the risk level, a pre-built level mapping table is searched to determine and execute the fund operation authority level corresponding to the fund outflow request of the blockchain account.

2. A bank funds settlement method based on blockchain according to claim 1, characterized in that: The receiving of a fund outflow request from a blockchain account and verifying the fund outflow request using a pre-deployed smart contract includes: Determining whether the fund outflow request meets the preset transaction rules through a smart contract, wherein the preset transaction rules include transaction amount limit, counterparty whitelist, transaction time window limit and transaction frequency limit; Confirm through a smart contract whether the number of nodes recognized by the fund outflow request in the blockchain network reaches a threshold, wherein the threshold is dynamically determined based on the transaction characteristics of the fund outflow request.

3. A bank funds settlement method based on blockchain according to claim 2, characterized in that: The determining whether the fund outflow request satisfies the preset transaction rules through the smart contract includes: a. Determine whether the transaction amount of the fund outflow request exceeds the set maximum transaction limit; If the transaction amount exceeds the maximum transaction limit, reject the fund outflow request; b. Check whether the recipient address of the fund outflow request is in the preset whitelist of the account; If the recipient address is not in the whitelist, the fund outflow request is rejected; c. Confirm whether the time point of the fund outflow request is within the set permitted trading time period; If the transaction time point is not within the permitted time period, the fund outflow request is rejected; d. Count the number of fund outflows from the account within the set time window; Determine whether the number of fund outflows from the blockchain account exceeds a set maximum frequency; If the number of fund outflows from the blockchain account exceeds the maximum frequency, the fund outflow request is rejected.

4. A bank funds settlement method based on blockchain according to claim 3, characterized in that: The step of confirming through a smart contract whether the number of nodes in the blockchain network that recognize the fund outflow request has reached a threshold includes: Dynamically determine a threshold value based on the transaction characteristics of the fund outflow request; Confirming whether the number of nodes in the blockchain network that have recognized the fund outflow request has reached the dynamically determined threshold; If the number of nodes does not exceed the threshold, rejecting the fund outflow request; If the number of nodes exceeds the threshold, proceed to the next step.

5. A bank funds settlement method based on blockchain according to claim 4, characterized in that: The method for dynamically determining the threshold value includes: Assign a weight to each transaction feature, wherein the transaction features include transaction amount A, user credit score C, transaction frequency F, transaction time T, and transaction location L; The comprehensive risk score of the fund outflow request is calculated based on the weight, wherein the calculation formula of the comprehensive risk score is: R=ω A ·A+ω C ·C+ω F ·F+ω T ·T+ω L ·L Among them, R represents the comprehensive risk score, ω A ,ω C ,ω F ,ω T and ω L They represent the weights of transaction amount A, user credit score C, transaction frequency F, transaction time T and transaction location L respectively, and their sum is 1; The threshold is dynamically determined according to the comprehensive risk score, wherein the formula for determination is: N=N0+k·R Wherein N0 represents a preset basic threshold, k represents an adjustment coefficient, and N represents a determined threshold.

6. A bank funds settlement method based on blockchain according to claim 5, characterized in that: The process of constructing the risk assessment model includes: Collect historical transaction data from the blockchain network; Using historical transaction data to train a risk assessment model to minimize prediction errors, wherein the risk assessment model uses an XGBoost algorithm; Deploy the trained risk assessment model based on the XGBoost algorithm to the blockchain network.

7. A bank funds settlement method based on blockchain according to claim 6, characterized in that: The dynamically adjusting the default risk level segment of the blockchain account based on the transaction characteristics of the fund outflow request to obtain the current risk level segment includes: According to the attribute of the blockchain account, searching for a default risk level segment of the blockchain account from a pre-built risk level segment mapping table, wherein the risk level segment mapping table includes default risk level segments corresponding to different attributes; According to the comprehensive risk score, the intervals in the default risk level segment are adjusted in sections to obtain the current risk level segment.

8. A bank funds settlement method based on blockchain according to claim 7, characterized in that: The risk grading is performed based on risk grading rules, wherein the risk grading rules include multiple risk levels and corresponding scoring ranges, and the risk levels include low risk, medium risk and high risk.

9. A bank funds settlement method based on blockchain according to claim 8, characterized in that: The searching a pre-built level mapping table based on the risk level, determining and executing the fund operation permission level corresponding to the fund outflow request of the blockchain account, includes: According to the risk level of the blockchain account, searching from a pre-built level mapping table to obtain the fund operation permission level corresponding to the risk level, wherein the level mapping table contains fund operation permission levels corresponding to different risk levels; According to the determined fund operation authority level, the operation corresponding to the fund operation authority level is performed.

10. A blockchain-based bank funds settlement system, used to execute a blockchain-based bank funds settlement method as claimed in any one of claims 1 to 9, characterized in that: The system comprises: A transaction verification module, used to receive a fund outflow request from a blockchain account and verify the fund outflow request using a pre-deployed smart contract; A risk assessment module, configured to input the transaction features associated with the fund outflow request into a pre-built risk assessment model to obtain a risk score after verification, wherein the risk assessment model is a machine learning model trained based on historical transaction data; A risk level adjustment module, configured to dynamically adjust a default risk level segment of the blockchain account based on the transaction characteristics of the fund outflow request to obtain a current risk level segment; A risk grading module, used to perform risk grading on the blockchain account in combination with the current risk level segment and the risk score to obtain the risk level of the blockchain account; The level verification module is used to search a pre-built level mapping table based on the risk level, determine the fund operation authority level corresponding to the fund outflow request of the blockchain account, and execute it.

Images