Dynamic decision-making method and device for payment routing, medium and product

Abstract

The invention provides a dynamic decision-making method and device for payment routing, a medium and a product, and belongs to the technical field of data processing, and the method comprises the steps: collecting payment information of a plurality of payment channels within a preset time length before the current time; determining a first weight of each payment channel based on the payment information, the first weight being related to a time decay factor, the time decay factor being related to an update period of the first weight; arranging the first weights of all payment channels to obtain a weight list; and in response to the received payment request, generating a routing policy of the payment request based on the weight list. According to the invention, the weight of each payment channel is adjusted based on the latest payment information of each payment channel and the time decay factor, so that the weight list of all payment channels is dynamically adjusted, the routing strategy is dynamically adjusted, the dynamic adaptive capacity of the payment routing scheme is improved, and the user experience is improved. The payment success rate and efficiency are improved, the transaction cost is reduced, and the stability of the payment system is guaranteed.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, and particularly relates to a dynamic decision method and device for payment routing, a medium and a product. BACKGROUND

[0002] Currently, a payment system usually relies on multiple payment channels, and a traditional payment routing scheme usually adopts a fixed priority or a manual configuration strategy. However, in actual business, such a payment routing scheme cannot adapt to changes in a real-time transaction environment, and may cause payment failure or a decrease in payment efficiency due to some failures of the payment channels (such as channel congestion during peak hours, sudden channel downtime, etc.). SUMMARY

[0003] The present application provides a dynamic decision method and device for payment routing, a medium and a product, and aims to solve the problem of lack of dynamic adjustment capability of a traditional payment routing scheme.

[0004] In a first aspect, the present application provides a dynamic decision method for payment routing, comprising: collecting payment information of multiple payment channels within a preset time length before a current time; determining a first weight of each payment channel based on the payment information, the first weight being related to a time decay factor, and the time decay factor being related to an update period of the first weight; arranging the first weights of all the payment channels to obtain a weight list; in response to receiving a payment request, generating a routing strategy of the payment request based on the weight list.

[0005] In an embodiment, the dynamic decision method for payment routing further comprises: during a payment process, dynamically adjusting the weight list based on real-time payment information.

[0006] In an embodiment, the first weight of each payment channel is determined based on the payment information, specifically comprising: determining multiple indicators of each payment channel based on the payment information; determining the first weight of the payment channel based on all the indicators of the payment channel.

[0007] In an embodiment, the multiple indicators of each payment channel are determined based on the payment information, specifically comprising: for each indicator, determining a basic data sequence of the indicator based on payment information related to the indicator; inputting the basic data sequence of each indicator into a corresponding autoregressive integrated moving average model to obtain an indicator output by the autoregressive integrated moving average model.

[0008] In an embodiment, the first weight of each payment channel is calculated based on all indicators of the payment channel, specifically including: The weighted sum of all indicators of the payment channel is taken as the first weight of the payment channel.

[0009] In an embodiment, the dynamic decision-making method of the payment routing further includes: In the payment process, the second weight of each indicator of each payment channel is dynamically adjusted based on real-time payment information.

[0010] In an embodiment, the calculation formula of the time decay factor is as follows: ; Wherein, is the base of natural logarithm; is a coefficient, which is a preset value; is the current time; is the adjustment time of the last first weight; represents the update period of the first weight; is the time decay factor.

[0011] In a second aspect, the present application also provides a dynamic decision-making device for payment routing, comprising: A receiving module is configured to collect payment information of a plurality of payment channels within a preset time length before the current time; A weight determination module is configured to determine the first weight of each payment channel based on the payment information, the first weight being related to a time decay factor, and the time decay factor being related to the update period of the first weight; A list obtaining module is configured to arrange the first weights of all payment channels to obtain a weight list; A routing allocation module is configured to generate a routing strategy of a payment request based on the weight list in response to receiving the payment request.

[0012] In a third aspect, the present application also provides an electronic device, which comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements any of the above-mentioned dynamic decision-making methods for payment routing when executing the computer program.

[0013] In a fourth aspect, the present application also provides a non-transitory computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement any of the above-mentioned dynamic decision-making methods for payment routing.

[0014] In a fifth aspect, the present application also provides a computer program product, which comprises a computer program, and the computer program is executed by a processor to implement any of the above-mentioned dynamic decision-making methods for payment routing. BRIEF DESCRIPTION OF DRAWINGS

[0015] In order to more clearly illustrate the technical solutions in the application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description are some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative effort on the basis of these drawings.

[0016] Figure 1 is one of the flow schematic diagrams of the dynamic decision method of payment routing provided by the present application; Figure 2 is the second flow schematic diagram of the dynamic decision method of payment routing provided by the present application; Figure 3 is one of the flow schematic diagrams of determining a plurality of indicators of each payment channel based on payment information provided by the present application; Figure 4 is one of the flow schematic diagrams of optimizing the second weight through reinforcement learning provided by the present application; Figure 5 is one of the structural schematic diagrams of the dynamic decision device of payment routing provided by the present application; Figure 6 is the structural schematic diagram of the electronic device provided by the present application. DETAILED DESCRIPTION

[0017] In order to make the purpose, technical solutions and advantages of the present application more clear, the technical solutions in the present application will be described clearly and completely in the following with reference to the drawings in the present application. Obviously, the described embodiments are some embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative effort belong to the scope of protection of the present application.

[0018] It should be noted that in the description of the present application, the terms "comprise", "include" or any other variants thereof are intended to cover non-exclusive inclusion, so that the process, method, article or device including a series of elements not only includes those elements, but also includes other elements not explicitly listed or inherent to such process, method, article or device. Without more limitation, the element defined by the statement "comprises a" does not exclude the presence of another identical element in the process, method, article or device including the element.

[0019] The terms "first", "second", etc. in this application are used to distinguish similar objects, and are not used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances, so that the embodiments of the present application can be implemented in an order other than those illustrated or described herein, and the objects distinguished by "first", "second", etc. are generally a class and do not limit the number of objects, for example, the first object can be one or more. In addition, "and / or" means at least one of the connected objects, and the character " / ", generally indicates that the front and rear associated objects are in a "or" relationship.

[0020] The following will be described in combination with Figures 1 to 6 The dynamic decision method of payment routing provided by the present application is described.

[0021] The dynamic decision method of payment routing provided by the embodiments of the present application is applicable to an insurance payment system, and is also applicable to other payment systems, such as a retail payment system.

[0022] It should be noted that the dynamic decision method of payment routing provided by the embodiments of the present application is implemented based on a dynamic decision device of payment routing. The dynamic decision method of payment routing can determine the weight of each payment channel based on multiple indexes, avoiding the limitations brought by determining the routing decision based on a single index. In addition, the weight of each payment channel can be dynamically adjusted based on multiple indexes of the payment channel, so as to dynamically adjust the routing strategy to adapt to the dynamically changing payment environment. Moreover, the embodiments of the present application adopt an automatic response mechanism of failure, avoiding the slow response speed, poor user experience and business discontinuity problems brought by subsequent manual intervention.

[0023] The dynamic decision method of payment routing is described by taking the dynamic decision device of payment routing as an execution subject.

[0024] Figure 1 is one of the flowcharts of the dynamic decision method of payment routing provided by the present application. Figure 2 is another flowchart of the dynamic decision method of payment routing provided by the present application.

[0025] In a possible implementation manner, as Figure 1 shown, the dynamic decision method of payment routing provided by the present application comprises: S110: Collect payment information of multiple payment channels within a preset time length before the current time.

[0026] The payment information includes the number of transactions, transaction results (success or failure), response time of each transaction, transaction amount, channel limit, transaction fee rate, fault alarm information and fault alarm level within the preset time length before the current time.

[0027] In one possible implementation, please combine Figure 2 During the user payment process, the data source (such as the payment platform) generates some raw data, such as service hardware indicators, payment gateway transaction records, and service logs. The data acquisition layer of the payment routing dynamic decision-making device collects data from the data source to obtain the aforementioned payment information.

[0028] One possible implementation involves using a monitoring system such as Prometheus to receive fault alarm information from each payment channel in real time and identify the fault level.

[0029] S120: Determine the first weight for each payment channel based on payment information. The first weight is related to a time decay factor, which in turn is related to the update cycle of the first weight.

[0030] In some existing payment routing schemes, the weights of each payment channel do not change with the time period. Such schemes do not take into account the impact of time on the performance of payment channels (for example, the payment success rate will fluctuate during peak hours during the day and off-peak hours at night), which causes the adjustment of the weights of payment channels to lag behind the changes in actual business scenarios.

[0031] Based on this consideration, in step S120, after analyzing the payment information of each payment channel within a preset time length before the current time, the comprehensive score of the payment channel is determined by combining the time decay factor, which serves as the first weight of the payment channel.

[0032] The time decay factor is used to reduce the influence of historical data on current decisions, emphasizing the importance of recent data. Specifically, the formula for calculating the time decay factor is as follows: (1); in, is the base of the natural logarithm; The coefficient is a preset value; The current time; This is the time since the last adjustment of the first weight; This indicates the update period for the first weight, which is the time interval for calculating the first weight. This represents the time decay factor. The following example illustrates the impact of the time decay factor on the first weight: A certain payment channel had a response time of 200ms 10 minutes ago (based on historical data, the response time metric is good), but in the last minute, it experienced 5 response timeouts (based on the latest data, the response time metric is poor). Comparison calculations: Determine the time decay factor based on historical data from 10 minutes ago: (2); Determine the time decay factor based on the latest data from 1 minute ago: (3).

[0033] As shown above, the time decay factor determined by the latest data is 221 times that determined by the historical data. The significant difference in the first weight of the payment channel determined based on these two time decay factors (see explanation below) ultimately has a significant impact on the routing strategy.

[0034] In addition, in step S120, the first weights of multiple payment channels are calculated in parallel, which can improve the efficiency of dynamic decision-making.

[0035] In one possible implementation, please combine Figure 2 The first weight of each payment channel is determined through a real-time computing engine and a decision service layer. Some basic data required by the decision service layer (such as time decay factors) is obtained from the storage system.

[0036] S130: Sort the first weights of all payment channels to obtain a weight list.

[0037] In one possible implementation, the first weights are sorted in descending order to obtain a weight list.

[0038] In one possible implementation, please combine Figure 2 The decision service layer ranks all payment channels by their first weight to obtain a weight list.

[0039] S140: In response to receiving a payment request, generate a routing strategy for the payment request based on the weight list.

[0040] In one possible implementation, a weighted random algorithm is used to determine the routing strategy. Specifically, firstly, the probability distribution of each payment channel is determined based on the first weight of all payment channels, thereby determining the selection range for each payment channel. Then, a random number is generated, and the payment channel corresponding to the selection range in which the random number falls is selected as the payment channel for the current payment request.

[0041] Weighted random algorithms can introduce a certain degree of randomness while maintaining the overall optimization direction, avoiding excessive concentration on a few channels, thereby improving the robustness of the system.

[0042] In another possible implementation, the payment channel with the highest weight in the weight list is selected to execute the payment, thereby determining the routing strategy. This decision-making method is more efficient and reliable.

[0043] In another possible implementation, a backup channel is automatically activated during peak periods to share a portion of the payment traffic (e.g., 30%). With equal weighting, each channel can handle its corresponding traffic, avoiding the pressure of concentrating traffic on a single payment channel and thus achieving peak shaving.

[0044] In one possible implementation, please combine Figure 2 The decision service layer generates a routing strategy for payment requests based on the weight list and sends it to the payment gateway corresponding to the payment channel in the routing strategy to execute the payment.

[0045] In one possible implementation, the decision service layer also records logs for each route switch (i.e., the situation of switching payment channels during the same payment process), including the switch time, switch reason, original payment channel, and new payment channel, for subsequent system optimization and troubleshooting.

[0046] This application's embodiments adjust the weights of each payment channel based on the latest payment information and time decay factors, thereby dynamically adjusting the weight list of all payment channels. This allows for dynamic adjustment of the routing strategy, improving the dynamic adaptability of the payment routing scheme, increasing payment success rate and efficiency, reducing transaction costs, and ensuring the stability of the payment system. Statistics show that the payment routing scheme based on this application can complete thousands of routing strategies per second, and system availability is increased from 99.5% to 99.95%.

[0047] Existing payment routing schemes select payment channels based solely on single-dimensional indicators such as success rate or transaction fee rate, lacking comprehensive evaluation of multiple factors, which leads to reduced payment efficiency.

[0048] In one possible implementation, step S120, determining the first weight of each payment channel based on payment information, specifically includes: S1201: Determine multiple metrics for each payment channel based on payment information.

[0049] In one possible implementation, multiple metrics include, but are not limited to: success rate metrics, response time metrics, available credit metrics, fee rate metrics, and fault alarm metrics.

[0050] In one possible implementation, please combine Figure 2 The real-time computing engine determines multiple indicators for each payment channel based on payment information, and then writes the indicators into the decision service layer.

[0051] S1202: Determine the first weight of each payment channel based on all indicators of each payment channel.

[0052] In one possible implementation, please combine Figure 3The decision service layer determines the first weight of each payment channel based on multiple indicators.

[0053] This application embodiment dynamically determines the weight of each payment channel by combining multiple dimensions such as success rate, response time, available credit, fee rate, and fault alarm, which better reflects the actual impact of payment channels on payments and facilitates more accurate routing strategies.

[0054] In one possible implementation, such as Figure 2 As shown, in step S1201, multiple indicators for each payment channel are determined based on payment information, specifically including: S310: For each indicator, determine the basic data sequence of the indicator based on the payment information associated with the indicator.

[0055] Specifically, the data stream within a preset time period before the current time is segmented for analysis, and an analysis result is obtained for each segment, thereby obtaining a sequence as the basic data sequence.

[0056] Based on the above-mentioned implementation examples of multiple indicators, the calculation methods for each indicator are as follows: a. The basic data sequence of the success rate indicator S: The basic data sequence of the success rate indicator S is a data sequence composed of payment success rates. The success rate indicator S is positively correlated with the payment success rate. An increase in the success rate of a single payment may lead to an increase in the success rate indicator S, thereby increasing the first weight of the payment channel (see the following explanation of step S1202).

[0057] When obtaining the payment success rate, a sliding window (e.g., window size of 1 minute) is used to segment transaction information from the same payment channel within a preset time length (e.g., 5 minutes, 10 minutes, or 1 hour) prior to the current time. The number of successful transactions and the total number of transactions are obtained within each sliding window. Then, the payment success rate is calculated for each sliding window. Payment success rate = number of successful transactions / total number of transactions (4).

[0058] One possible implementation is to use the Flink processing engine to perform interactive window computation.

[0059] b. The underlying data sequence for the response time metric T: The basic data sequence of the response time index T is a data sequence composed of response times. The response time index T is positively correlated with the response time. An increase in a single response time data may lead to an increase in the response time index T, thereby reducing the first weight of the payment channel (see the following explanation of step S1202).

[0060] When obtaining the basic data sequence for response time metrics, the transaction information of the same payment channel within a preset time length before the current time is first segmented using a sliding window. Then, the response time of each sliding window is determined based on the transaction information within that sliding window.

[0061] In one possible implementation, to determine the response time of each sliding window, the transaction amount and original response time of each payment within each sliding window are first obtained. Then, the weighted average time of the sliding window is calculated as the response time of that window. (5); in, The number of payments within the sliding window; the transaction amount percentage is the percentage of each payment's transaction amount within the total transaction amount of the sliding window.

[0062] In another possible implementation, the average response time of all response times within the sliding window, or the maximum response time among all response times within the sliding window, is used as the response time of the sliding window.

[0063] c. Basic data sequence of available credit limit indicator R: The basic data sequence of the available credit limit indicator R is a data sequence composed of the percentage of remaining credit limit. An increase in the percentage of a single remaining credit limit may lead to an increase in the available credit limit indicator R, thereby increasing the first weight of the payment channel (see the following explanation of step S1202).

[0064] When obtaining the basic data sequence of the available credit limit indicator R, the transaction information of the same payment channel within a preset time period before the current time is first segmented using a sliding window. Then, the used credit limit and channel limit are obtained from each sliding window, and the remaining credit limit percentage is determined based on the used credit limit and channel limit within that sliding window. Remaining quota percentage = (channel limit - used quota) / channel limit (6).

[0065] d. The underlying data sequence for the rate indicator F: The fee rate indicator F is negatively correlated with the commission rate.

[0066] In one possible implementation, the basic data sequence of the fee rate indicator F for each payment channel is a data sequence consisting of the fee rate coefficients for each payment in the same payment channel within a preset time period prior to the current time. Fee rate coefficient = 1 / (1 + handling fee rate) (7).

[0067] Based on a tiered fee strategy, lower transaction fees result in lower payment costs and higher fee coefficients. When determining the fee index F based on a data sequence composed of fee coefficients, an increase in a single fee coefficient may lead to an increase in the fee index F, thereby increasing the first weight of the payment channel (see the explanation of step S1202 below). In other words, payment channels with lower transaction fees may be allocated more payment traffic, thus reducing costs.

[0068] Based on the above, according to the daily transaction limit of the channel or the tiered fee rate, when making payment routing decisions (step S140), large transaction orders will be automatically diverted to low-cost channels to save daily costs.

[0069] e. Basic data sequence of fault alarm indicator H: The fault alarm indicator H is a punitive indicator for fault alarms.

[0070] In one possible implementation, the basic data sequence of the fault alarm index H for each payment channel is the level of each fault alarm of the same payment channel within a preset time length before the current time (e.g., no alarm, minor, medium, severe).

[0071] An increase in the level of a fault alarm (i.e., a more severe fault) will lead to a decrease in the fault alarm index H, and the first weight of the payment channel will decrease accordingly (see the following explanation of step S1202).

[0072] S320: Input the basic data sequence of each indicator into the corresponding Autoregressive Integrated Moving Average (ARIMA) model to obtain the indicators output by the ARIMA model, such as the success rate indicator S, response time indicator T, available credit indicator R, fee rate indicator F, and fault alarm indicator H mentioned above.

[0073] Specifically, each indicator in each payment channel has an ARIMA model, which is used to infer the value of the indicator in the next time step of the payment channel based on the basic data sequence of the indicator in that payment channel.

[0074] The embodiments of this application predict the indicators based on the basic data sequences of each indicator using the ARIMA model. Compared with obtaining the indicators directly from the original data, the numerical values ​​of the indicators are more accurate.

[0075] In one possible implementation, step S1202 involves calculating the first weight of each payment channel based on all indicators of that payment channel, specifically including: The weighted sum of all metrics for the payment channel is used as the first weight for the payment channel: (8); in, , , , , This is the second weight for each indicator.

[0076] This application embodiment determines the first weight of each payment channel by weighting and summing dynamically changing multi-dimensional indicators. This scheme integrates real-time transaction data and system monitoring indicators to achieve adaptive adjustment of payment channel weights.

[0077] In one possible implementation, the dynamic decision-making method for payment routing also includes: The second weight is optimized periodically or irregularly.

[0078] Specifically, please combine Figure 4 In the initial stage of dynamic payment routing decision-making, a cache preloading mechanism is used to retrieve and load the first weights of each payment channel from the storage system. This allows for payment routing decisions to be made using the initialized payment channel weights in the early stages, reducing computational latency. Furthermore, in the initial stage of dynamic payment routing decision-making, the second weights of each indicator in each payment channel (such as those mentioned above) are retrieved from the storage system. , , , , Then initialize the second weight.

[0079] As the dynamic payment routing process continues, the system can update various indicators based on the real-time collected payment information, thereby updating the primary weight of each payment channel.

[0080] Based on the above, since the environment of each payment channel is constantly changing, the second weight also needs to be dynamically changed as the payment environment changes. Therefore, the second weight needs to be optimized regularly or irregularly.

[0081] In one possible implementation, such as Figure 5 As shown, the second weight is optimized through reinforcement learning, specifically including the following steps: S410: Environment Modeling and State Space Definition.

[0082] Define the state space as And initialize the second weight (e.g., the one mentioned above). , , , , The initial second weight is preset based on historical performance, business priority, etc.

[0083] S420: Set the reward function .

[0084] S430: Configure the policy network architecture.

[0085] In one possible implementation, a hierarchical strategy network design is adopted, and iteration is achieved by adjusting the subnetworks of the hierarchical strategy network through parameter adjustments.

[0086] S440: Model training.

[0087] Specifically, transaction records for each payment are obtained based on the collected payment information, such as [timestamp, channel ID, delay, cost, state vector], as samples.

[0088] In each iteration of the training process, a batch of samples is sampled. Based on the sampling results, a model update strategy is determined, and a reward function is determined based on the execution result of the update strategy. Subsequently, a loss function is determined based on the reward function. The network parameters are updated through backpropagation and gradient descent, thus achieving iteration. During the iteration process, the learning rate is adaptively adjusted. Finally, the optimized second weights (such as those mentioned above) are obtained. , , , , ).

[0089] S450: Verify the optimized second weight.

[0090] In one possible implementation, offline simulation testing and / or online A / B testing can be used for verification.

[0091] This application embodiment dynamically adjusts the second weight through reinforcement learning, so that the weight calculation of payment routing dynamically adapts to changes in business scenarios.

[0092] In one possible implementation, the dynamic decision-making method for payment routing also includes: During the payment process, the second weight of each indicator for each payment channel is dynamically adjusted based on real-time payment information.

[0093] In one possible implementation, the second weight of each indicator for each payment channel is dynamically adjusted based on real-time payment information, specifically including at least one of the following: During the payment process, if a fault alarm is triggered in the first payment channel, the second weight of the fault alarm indicator H of the first payment channel will be adjusted. Set to 0 and trigger an emergency downgrade of the first payment channel (e.g., reduce the first weight of the first payment channel), while determining the routing strategy based on the updated weight list; During the payment process, if the response time of the second payment channel exceeds the response time threshold, an exponential backoff algorithm is used to reduce the response time index of the second payment channel. Second weight At the same time, the routing strategy is determined based on the updated weight list; During the payment process, an exponential decay model is used to adjust the second weight of the response time indicator T. If the timeout rate does not exceed the threshold, the second weight... No decay (i.e., equal to 1); second weight if the timeout rate exceeds the threshold. Decrease (e.g., 0.9, 0.8, etc.); the more the timeout rate exceeds the threshold, the higher the weight. The greater the attenuation; During the payment process, an exponential decay model is used to adjust the second weight of the available credit limit indicator R. That is, as the available credit limit indicator R decreases, the second weight... The decrease is exponential, meaning that the second weight of payment channels with low available funds decreases exponentially. During the payment process, the second weight of the fee rate indicator F is adjusted by combining a tiered fee rate strategy. From the above relationship between the fee rate indicator F and the transaction fee rate, it can be seen that the lower the transaction fee rate, the lower the payment cost, and the higher the fee rate indicator F. In the payment process, the second weight... It increases as the rate indicator F increases.

[0094] The embodiments of this application dynamically adjust the second weight of the indicators that are abnormal based on the real-time payment status, so as to dynamically adjust the weight list, adjust the routing strategy in a timely manner, and improve dynamic adaptability.

[0095] Existing failover mechanisms typically rely on manual intervention after the fact, resulting in slow fault response and failover delays of several minutes, impacting user experience and business continuity, and failing to isolate faults in a timely manner.

[0096] Based on the above considerations, in one possible implementation, the dynamic decision-making method for payment routing also includes: During the payment process, the weight list is dynamically adjusted based on real-time payment information.

[0097] In one possible implementation, during the payment process, the weight list is dynamically adjusted based on real-time payment information, specifically including at least one of the following: During the payment process, if a payment channel meets preset conditions, the payment channel will be removed from the weight list or its priority will be reduced. The preset conditions are that the payment channel's first weight is lower than a weight threshold, or the payment channel's fault alarm level is higher than a preset level.

[0098] Based on the above, in one instance, when the payment channel with the highest weight fails due to momentary reasons, the system can automatically switch to the payment channel with the second highest weight, thus achieving failover.

[0099] This application embodiment triggers the switching of payment channels in real time by determining a threshold, reducing manual intervention and realizing real-time fault detection and real-time channel switching. The fault switching response time is shortened from minutes to milliseconds, ensuring business continuity.

[0100] Based on the above, in this application, manual intervention is only performed in extreme failure scenarios (such as a complete channel outage) or major strategy changes (such as fee tier adjustments), reducing the intervention frequency from 1-2 times per hour in the traditional approach to less than once per day. Statistics show that the payment routing scheme based on this application reduces the manual operation error rate by 95%.

[0101] One possible implementation is to use a hybrid deployment model for payment routing decisions: for example, 80% of payment traffic uses a dynamic strategy, while 20% retains a fixed strategy as a safety net to ensure business reliability.

[0102] Based on the above, this application also provides a dynamic decision-making device for payment routing. The dynamic decision-making device for payment routing and the aforementioned dynamic decision-making method for payment routing can be referred to and correspond to each other.

[0103] As an example, such as Figure 6 As shown, the dynamic decision-making device for payment routing includes: The receiving module 510 is used to collect payment information from multiple payment channels within a preset time period prior to the current time; The weight determination module 520 is used to determine the first weight of each payment channel based on payment information. The first weight is related to the time decay factor, which is related to the update cycle of the first weight. The list retrieval module 530 is used to sort all payment channels by their first weight and obtain a weight list; The routing allocation module 540 is used to generate a routing strategy for the payment request based on the weight list in response to the receipt of the payment request.

[0104] The embodiments of this application adjust the weight of each payment channel based on the latest payment information and time decay factor, thereby dynamically adjusting the weight list of all payment channels. This enables the routing strategy to be dynamically adjusted, improving the dynamic adaptability of the payment routing scheme, increasing payment success rate and efficiency, reducing transaction costs, and ensuring the stability of the payment system.

[0105] In one possible implementation, the dynamic decision-making device for payment routing further includes a weight list adjustment module 550, which is used to dynamically adjust the weight list based on real-time payment information during the payment process.

[0106] This application embodiment triggers the switching of payment channels in real time by determining a threshold, reducing manual intervention and realizing real-time fault detection and real-time channel switching. The fault switching response time is shortened from minutes to milliseconds, ensuring business continuity.

[0107] In one possible implementation, the weight determination module 520 is specifically used for: Multiple metrics for each payment channel are determined based on payment information; The first weight of each payment channel is determined based on all indicators of each payment channel.

[0108] This application embodiment dynamically determines the weight of each payment channel by combining multiple dimensions such as success rate, response time, available credit, fee rate, and fault alarm, which better reflects the actual impact of payment channels on payments and facilitates more accurate routing strategies.

[0109] In one possible implementation, the dynamic decision-making device for payment routing further includes a weight adjustment module 560, which is used to dynamically adjust the second weight of each indicator of each payment channel based on real-time payment information during the payment process.

[0110] The embodiments of this application dynamically adjust the second weight of the indicators that are abnormal based on the real-time payment status, so as to dynamically adjust the weight list, adjust the routing strategy in a timely manner, and improve dynamic adaptability.

[0111] Figure 6 This is a schematic diagram of the structure of the electronic device provided in this application, such as... ​ As shown, the electronic device may include: a processor 610, a communications interface 620, a memory 630, and a communication bus 640, wherein the processor 610, the communications interface 620, and the memory 630 communicate with each other via the communication bus 640. The processor 610 can call logical instructions in the memory 630 to execute a dynamic decision-making method for payment routing, the method including: Collect payment information from multiple payment channels within a preset time period prior to the current time; The first weight of each payment channel is determined based on payment information. The first weight is related to the time decay factor, which is related to the update cycle of the first weight. Arrange all payment channels by their first weight to obtain a weight list; In response to receiving a payment request, a routing strategy for the payment request is generated based on the weight list.

[0112] Furthermore, the logical instructions in the aforementioned memory 630 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0113] On the other hand, this application also provides a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium. The computer program includes program instructions, and when the program instructions are executed by a computer, the computer is able to execute the dynamic decision-making method for payment routing provided in the above embodiments. The method includes: Collect payment information from multiple payment channels within a preset time period prior to the current time; The first weight of each payment channel is determined based on payment information. The first weight is related to the time decay factor, which is related to the update cycle of the first weight. Arrange all payment channels by their first weight to obtain a weight list; In response to receiving a payment request, a routing strategy for the payment request is generated based on the weight list.

[0114] In another aspect, this application also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements a dynamic decision-making method for payment routing provided in the above embodiments, the method comprising: Collect payment information from multiple payment channels within a preset time period prior to the current time; The first weight of each payment channel is determined based on payment information. The first weight is related to the time decay factor, which is related to the update cycle of the first weight. Arrange all payment channels by their first weight to obtain a weight list; In response to receiving a payment request, a routing strategy for the payment request is generated based on the weight list.

[0115] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0116] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of various embodiments or some parts of embodiments.

[0117] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A dynamic decision-making method for payment routing, characterized in that, include: Collect payment information from multiple payment channels within a preset time period prior to the current time; A first weight is determined for each payment channel based on the payment information. The first weight is related to a time decay factor, which is related to the update cycle of the first weight. Arrange all payment channels by their first weight to obtain a weight list; In response to receiving a payment request, a routing strategy for the payment request is generated based on the weight list.

2. The dynamic decision-making method for payment routing according to claim 1, characterized in that, Also includes: During the payment process, the weight list is dynamically adjusted based on real-time payment information.

3. The dynamic decision-making method for payment routing according to claim 1, characterized in that, Determining a first weight for each payment channel based on the payment information specifically includes: Based on the payment information, multiple indicators are determined for each payment channel; The first weight of each payment channel is determined based on all the indicators of each payment channel.

4. The dynamic decision-making method for payment routing according to claim 3, characterized in that, Based on the payment information, multiple indicators are determined for each payment channel, specifically including: For each indicator, a basic data sequence for that indicator is determined based on payment information associated with that indicator; The basic data sequence of each indicator is input into the corresponding autoregressive integral moving average model to obtain the indicator output by the autoregressive integral moving average model.

5. The dynamic decision-making method for payment routing according to claim 3, characterized in that, The first weight of each payment channel is calculated based on all indicators of each payment channel, specifically including: The weighted sum of all indicators of the payment channel is used as the first weight of the payment channel.

6. The dynamic decision-making method for payment routing according to claim 5, characterized in that, Also includes: During the payment process, the second weight of each indicator for each payment channel is dynamically adjusted based on real-time payment information.

7. The dynamic decision-making method for payment routing according to claim 1, characterized in that, The formula for calculating the time decay factor is as follows: ； in, is the base of the natural logarithm; The coefficient is a preset value; The current time; The time since the last adjustment of the first weight; This indicates the update period for the first weight; This is the time decay factor.

8. A dynamic decision-making device for payment routing, characterized in that, include: The receiving module is used to collect payment information from multiple payment channels within a preset time period prior to the current time; The weight determination module is used to determine a first weight for each payment channel based on the payment information. The first weight is related to a time decay factor, and the time decay factor is related to the update cycle of the first weight. The list retrieval module is used to sort all payment channels by their first weight and obtain a weight list; The routing allocation module is used to generate a routing strategy for the payment request based on the weight list in response to receiving the payment request.

9. A non-transitory computer-readable storage medium, wherein a computer program is stored on the non-transitory computer-readable storage medium, characterized in that, When the computer program is executed by a processor, it implements the dynamic decision-making method for payment routing as described in any one of claims 1 to 7.

10. A computer program product, the computer program product comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the dynamic decision-making method for payment routing as described in any one of claims 1 to 7.

Images