Driver fund sub-account method and system based on four-element verification and risk control

Abstract

The application discloses a driver fund sub-account method and system based on four-element verification and risk control, and belongs to the technical field of network freight fund settlement and anti-fraud. The method comprises the following steps: obtaining four identity elements in a driver withdrawal request and calling an identity verification channel to perform consistency verification, and refusing when no configuration or failure; real-time aggregation of withdrawable balance according to the driver dimension, and refusing when exceeding; real-time aggregation of withdrawal behavior statistics, execution of local rules for preliminary determination and double-channel cross verification, and refusal when hitting or determining interception; generation of a withdrawal order subject to a unique key constraint of a database; state machine flow transfer, payment with a withdrawal order number as a power identifier and use of conditional update to prevent repeated payment; verification, risk control determination and state change writing into an only-increase audit log.

Description

Technical Field

[0001] This invention relates to the field of fund settlement and anti-fraud technology for online freight platforms, and in particular to a driver fund distribution method and system based on four-factor verification and risk control. Background Technology

[0002] In online freight platforms, after drivers (usually individual transporters) complete their transport transactions, the platform distributes funds based on settlement statements, and drivers then initiate withdrawals of their earnings. Existing withdrawal schemes have the following drawbacks: First, withdrawals are only possible by linking a bank card, and the applicant's name, ID number, bank card number, and mobile phone number are not verified for consistency, posing a risk of impersonation and fund theft. Second, there is a lack of real-time identification mechanisms for abnormal behaviors such as high-frequency withdrawals, unusually large withdrawals, withdrawals over multiple consecutive days, and daily cumulative withdrawals exceeding limits, making it difficult to promptly detect risks such as funds being deposited and withdrawn immediately or misappropriated. Third, the status of withdrawal applications, reviews, payments, receipts, and rejections is not constrained by state machines, potentially leading to duplicate payments and unauthorized operations. Fourth, the payment process lacks idempotency mechanisms and concurrency control; network retries or concurrent requests may result in duplicate payments for the same withdrawal order, causing financial losses. Fifth, verification records, risk control judgments, and status changes lack auditing, failing to meet regulatory verification and anti-money laundering record retention requirements. Technical problems to be solved

[0003] To address the shortcomings of existing technologies, the technical problem this invention aims to solve is: how to achieve automatic verification of the identity of drivers withdrawing cash on online freight platforms, real-time identification and interception of abnormal withdrawal behavior, strict control over withdrawal status, effective prevention of duplicate payments, and full-process auditability. Technical solution

[0004] To solve the above-mentioned technical problems, the present invention provides a method for driver fund distribution, comprising the following steps:

[0005] Step S1, Four-Factor Verification: Respond to the driver's withdrawal request, obtain the driver's four identity factors (name, ID number, bank card number, and mobile phone number), and call the identity verification channel to perform consistency verification of the four factors; if the verification channel is not configured, reject the withdrawal request; after the verification is successful, write the verification record to the incremental audit table.

[0006] Step S2, Balance Verification: Aggregate the withdrawable balance in real time by driver (the total amount actually received by drivers whose earnings have been confirmed minus the total amount withdrawn and in transit by the driver). If the requested withdrawal amount is greater than the withdrawable balance, the application will be rejected.

[0007] Step S3, Anti-fraud Risk Control Guard: Aggregate the driver's withdrawal behavior statistics in real time (number of withdrawals in the recent time window, cumulative withdrawal amount on the day, number of consecutive withdrawal days, and single withdrawal amount), and perform local rule channel judgment and dual-channel cross-validation; if the local main rule channel hits, it will be rejected and transferred to manual review; if the cross-validation judgment is blocked, it will be rejected; if the verification is successful, a withdrawal order constrained by the database unique key will be generated and the initial state will be created.

[0008] Step S4, State Machine Transition: The withdrawal order transitions according to the preset state machine; the payment step calls the payment channel to execute the transfer, using the withdrawal order number as an idempotent identifier for reuse, and adopts conditional updates (only the audit status can be set to the payment status) to prevent concurrent duplicate payments; payment is rejected if the payment channel is not configured.

[0009] Step S5, full-process audit: Write the four-factor verification records, risk control judgment results, and changes in the status of each withdrawal slip into the incremental audit log, and provide query and anonymized display by driver, time, and status.

[0010] Furthermore, the four-factor verification includes: throwing an exception and rejecting the withdrawal application when the verification channel is not configured or the verification call fails, and prohibiting false approval.

[0011] Furthermore, the anti-fraud determination includes: a single withdrawal amount exceeding a first threshold, the number of withdrawals within a recent time window exceeding a second threshold, the cumulative withdrawal amount on a given day exceeding a third threshold, and the number of consecutive withdrawal days exceeding a fourth threshold, all of which are considered high-risk.

[0012] Furthermore, the anti-fraud determination also includes: first determining the main channel based on the result of the local rule channel, and blocking if the main channel hits; then performing cross-validation between the main channel and the complementary channel, and rejecting if the cross-validation result is blocking.

[0013] Furthermore, the state machine includes a sequential flow of "application → review → payment → receipt" and a flow of "application → rejection". Each flow verifies the current status, and illegal migrations are rejected.

[0014] Furthermore, the payment steps include: using the withdrawal order number as an idempotent identifier for the payment channel; and using a conditional update of "only the audited status can be set to the payment status" to prevent concurrent duplicate payments.

[0015] Furthermore, the withdrawable balance is aggregated by driver dimension, which includes the total amount of the driver's confirmed share of earnings minus the total amount of the driver's withdrawn and in-transit withdrawals.

[0016] Furthermore, in the aforementioned audit log, the ID card number and bank card number are displayed and returned in an anonymized form. Beneficial effects

[0017] Compared with the prior art, the present invention has the following beneficial effects:

[0018] (1) Real-time verification of the four elements is prohibited when the verification channel is not configured, which can prevent fraudulent withdrawals;

[0019] (2) By intercepting through local rules channels and cross-verifying through dual channels, abnormal behaviors such as high-frequency, large-amount, and continuous withdrawals can be detected in a timely manner, and the risk control of funds entering and leaving the anti-money laundering fund is connected.

[0020] (3) Through the unique key of the withdrawal order number database, the idempotent reuse of the payment channel, and the triple mechanism of condition update, duplicate payments caused by concurrent requests or network retries can be prevented.

[0021] (4) Through the valid state machine transition check, unauthorized and illegal state transitions cannot be executed;

[0022] (5) Verification, risk control, and status changes are all recorded and sensitive information is desensitized to meet the requirements of regulatory verification and anti-money laundering record retention. Attached Figure Description

[0023] Figure 1 This is a schematic diagram of the system module structure of the present invention;

[0024] Figure 2 This is a flowchart illustrating the overall process of the cash withdrawal and revenue sharing method of this invention.

[0025] Figure 3 This is a schematic diagram illustrating the single-state machine transition of the present invention;

[0026] Figure 4 This is a flowchart of the anti-fraud risk control guard determination process of the present invention. Detailed Implementation

[0027] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments. I. System Overall Architecture

[0028] like Figure 1 As shown, the system of this invention includes: a four-factor verification module, a revenue sharing plan module, a withdrawal application module, a risk control guard module, a state machine module, a payment idempotency module, a reconciliation module, and an audit module. Each module interacts with a database, which contains a four-factor verification table, a revenue sharing plan table, and a withdrawal order table (including a unique withdrawal order number constraint).

[0029] The four-factor authentication module is configuration-driven (verification channel address and key). It calls the real identity verification channel via HTTP. The request body includes name, ID number, bank card number, mobile phone number and channel key. It parses the pass flag and prompt information in the response. If the channel is not configured, it throws an exception and rejects the withdrawal request. II. Withdrawal and Revenue Sharing Process

[0030] like Figure 2 As shown, the method includes: receiving driver withdrawal requests → four-factor verification (reject if not configured / failure) → storing verification records → balance verification (reject if exceeding limit) → aggregation of behavioral statistics → local rule channel determination (reject if hit and transfer to manual review) → dual-channel cross-validation (BLOCK rejection) → generating a unique withdrawal order number and creating an "application" status → review → payment (idempotent order number + condition update) → receipt → final status; the entire process is logged in an audit log.

[0031] The unique withdrawal order number consists of a date prefix and a random string (such as WD+yyyyMMdd+16-bit random string), and is constrained by the database unique key. When multiple requests are submitted concurrently, unique key conflicts are blocked. III. State Machine

[0032] like Figure 3 As shown, withdrawal orders flow in the order of "Applied → Approved → Paid → Receipt," or directly from "Applied" to the final state of "Rejected." Each flow verifies the current status; illegal migrations trigger a business exception, and the status is not recorded in the database. The payment step uses the withdrawal order number as the idempotent identifier for the payment channel (the channel batch number reuses the withdrawal order number), and employs a conditional update of "only the approved status can be set to the paid status" to prevent concurrent duplicate payments; payments are rejected if the payment channel is not configured (fail-fast). IV. Anti-fraud risk control

[0033] like Figure 4 As shown, the risk control guard includes four rules: a single withdrawal reaching or exceeding the first threshold (e.g., 50,000 yuan); the number of withdrawals within a recent time window (e.g., 1 hour) reaching or exceeding the second threshold (e.g., 3 times); the cumulative withdrawal amount for the day reaching or exceeding the third threshold (e.g., 100,000 yuan); and the number of consecutive withdrawal days reaching or exceeding the fourth threshold (e.g., 3 days). The local master rule channel (Channel A) makes the initial judgment; if a match is found, the withdrawal is rejected and transferred to manual review. Subsequently, cross-validation is performed between Channel A and the complementary rule channel (Channel B); if the cross-validation result is a block, the withdrawal is rejected. Statistics are aggregated from the database in real time with each withdrawal request. V. Auditing and Desensitization

[0034] The four-factor verification records, risk control judgment results, and withdrawal order status changes are all written to the INSERT-only audit log, with no update or deletion entry; when querying the list, the ID number and bank card number are displayed in an anonymized form (retaining the last four digits) to ensure personal information security.

[0035] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for distributing driver funds, characterized in that, include: The system retrieves the four key identity elements from driver withdrawal requests, calls the identity verification channel to perform consistency verification of these elements, and rejects withdrawal requests if the verification channel is not configured or verification fails. It also aggregates available withdrawal balances by driver in real time, rejecting requests if the requested withdrawal amount exceeds the available balance. Furthermore, it aggregates driver withdrawal behavior statistics in real time to perform anti-fraud checks, rejecting requests if they hit risk control rules or are determined to be blocked through cross-validation. The system generates withdrawal orders constrained by unique database keys and creates an initial state. It then transitions the withdrawal order state according to a preset state machine, using the withdrawal order number as an idempotent identifier and employing conditional updates to prevent duplicate payments. Finally, it writes verification records, risk control judgments, and state changes to an increment-only audit log.

2. The driver fund distribution method according to claim 1, characterized in that, The four-factor consistency verification includes: throwing an exception and rejecting the withdrawal application when the identity verification channel is not configured or the verification call fails, and prohibiting false approval.

3. The driver fund distribution method according to claim 1, characterized in that, The anti-fraud determination includes: when a single withdrawal amount exceeds the first threshold, or the number of withdrawals within a recent time window exceeds the second threshold, or the cumulative withdrawal amount on a given day exceeds the third threshold, or the number of consecutive withdrawal days exceeds the fourth threshold, it is determined to be high risk.

4. The driver fund distribution method according to claim 1, characterized in that, The anti-fraud determination also includes: first determining the main channel based on the results of the local rule channel, and blocking if the main channel hits; then performing cross-validation between the main channel and the complementary channel, and rejecting if the cross-validation result is blocking.

5. The driver fund distribution method according to claim 1, characterized in that, The preset state machine includes a sequential flow of "application → approval → payment → receipt" and a flow of "application → rejection". Each flow verifies the current status, and illegal migrations are rejected.

6. The driver fund distribution method according to claim 1, characterized in that, The payment process includes: using the withdrawal order number as the idempotent identifier of the payment channel; using the conditional update of "only the audit status can be set to payment status" to prevent concurrent duplicate payments; and rejecting payments if the payment channel is not configured.

7. The driver fund distribution method according to claim 1, characterized in that, The withdrawable balance is aggregated by driver, representing the total amount actually received by the driver who has confirmed the split, minus the total amount withdrawn and in transit by the driver.

8. The driver fund distribution method according to claim 1, characterized in that, In the audit log, the ID card number and bank card number are displayed and returned in an anonymized form.

Images