Complex service processing method and system based on distributed intelligent agent system

Abstract

The embodiment of the invention provides a complex business processing method and system based on a distributed intelligent agent system, and the method comprises the steps: obtaining business data and business processing demands, carrying out the analysis processing of the business processing demands, and obtaining a plurality of subtasks, performing type allocation on each sub-task according to the attribute of the service data to obtain a corresponding agent type, determining a first distributed agent node based on a load resource prediction model and a load demand associated with the sub-task of the single-step agent, and processing the sub-task of the single-step agent to obtain a first result, according to the load rate and the communication delay time of each unselected distributed agent node, screening out a plurality of to-be-selected node combinations of the sub-tasks of the sequential agent, and determining a second node combination based on the matching degree among the to-be-selected nodes in each to-be-selected node combination for processing the sub-tasks of the sequential agent to obtain a second result, and fusing the first result and the second result to obtain a final result. According to the scheme, efficient processing of the business data can be realized.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of electronic digital data processing, and in particular, to a complex service processing method and system based on a distributed intelligent agent system. Background Art

[0002] With the rapid evolution of artificial intelligence technology, the complexity of its business processing tasks in fields such as finance, healthcare, and intelligent manufacturing is experiencing exponential growth. Current artificial intelligence systems can effectively handle simple business scenarios with high standardization and relatively single processes through lightweight model architectures and distributed computing frameworks. A distributed artificial intelligence system consists of multiple distributed agent nodes, each of which has autonomous decision-making capabilities, local resources (computing / storage / communication), and task execution logic, and collaborates with other nodes through a decentralized communication protocol to achieve complex business goals.

[0003] In related technologies, for single or relatively simple business processing, the efficiency can basically meet user needs and the cost is controllable. For example, tasks such as image classification and text translation can achieve millisecond-level responses through a distributed computing framework and the hardware cost is controllable. However, the processing efficiency for complex business problems is low and cannot meet the processing requirements. Summary of the Invention The embodiments of the present application provide a complex service processing method and system based on a distributed intelligent agent system, which solve the problem that the processing efficiency for complex business problems in the prior art is low and cannot meet the processing requirements, can achieve efficient processing of business data, and while ensuring the processing efficiency, reduce the processing cost.

[0004] In a first aspect, the embodiments of the present application provide a complex service processing method based on a distributed intelligent agent system, including: Obtain the business data to be processed and the corresponding business processing requirements, perform parsing processing on the business processing requirements to obtain multiple subtasks, and allocate an agent type to each of the subtasks according to the attributes of the business data to obtain the corresponding agent type, where the agent type includes a single-step agent type and a sequential agent type; Determine a first distributed agent node among multiple distributed agent nodes based on the load resource prediction model and the load requirements associated with the subtasks of the single-step agent type, and send the subtasks of the single-step agent type to the first distributed agent node for processing to obtain a first processing result; Screen out multiple candidate agent node combinations for the subtasks of the sequential agent type according to the load rates and communication delay times of each of the unselected distributed agent nodes, and determine a second agent node combination based on the matching degree between each candidate agent node in each of the candidate agent node combinations; Send the subtasks of the sequential proxy type to the second proxy node combination for processing to obtain a second processing result, and fuse the first processing result and the second processing result to obtain the final processing result of the service data.

[0005] Optionally, determining the first distributed proxy node among multiple distributed proxy nodes based on the load resource prediction model and the load requirements associated with the subtasks of the single-step proxy type includes: Input the current comprehensive load information and the time period of multiple distributed proxy nodes into the load resource prediction model to obtain the load resource prediction information of each distributed proxy node; Determine the first distributed proxy node for the subtasks of the corresponding single-step proxy type according to the load resource prediction information of each distributed proxy node and the load requirements.

[0006] Optionally, determining the first distributed proxy node for the subtasks of the corresponding single-step proxy type according to the load resource prediction information of each distributed proxy node and the load requirements includes: In the case where there are multiple subtasks of the single-step proxy type, compare the load requirements associated with each subtask of the single-step proxy type, and perform priority sorting according to the comparison result to obtain the allocation priorities of each subtask of the single-step proxy type; Compare the load requirements of the subtask with the lowest allocation priority with the load resource prediction information of each distributed proxy node to obtain an optional proxy node sequence combination, and determine the first distributed proxy node for each subtask of the single-step proxy type according to the allocation priority and the optional proxy node sequence combination.

[0007] Optionally, the subtasks of the sequential proxy type include multiple minimum division tasks that are sequentially executed. Screening out multiple candidate proxy node combinations for the subtasks of the sequential proxy type according to the load rates and communication delay times of the unselected distributed proxy nodes includes: Screen out multiple candidate proxy nodes with a load rate lower than the first threshold and a communication delay time lower than the second threshold from the unselected distributed proxy nodes, and perform combination processing on the multiple candidate proxy nodes according to the number of the minimum division tasks to obtain multiple candidate proxy node combinations.

[0008] Optionally, determining the second proxy node combination based on the matching degree between the candidate proxy nodes in each candidate proxy node combination includes: Arrange each of the candidate proxy nodes in each of the candidate proxy node combinations in sequence to obtain a corresponding plurality of sequential proxy node sequences; Determine a second proxy node combination based on the matching degrees between the candidate proxy nodes and the plurality of sequential proxy node sequences of each of the candidate proxy node combinations.

[0009] Optionally, the determining the second proxy node combination based on the matching degrees between the candidate proxy nodes and the plurality of sequential proxy node sequences of each of the candidate proxy node combinations includes: Superimpose the matching degrees between the candidate proxy nodes connected in sequence in the plurality of sequential proxy node sequences of the candidate proxy node combination to obtain a comprehensive matching degree, and determine the sequential proxy node sequence with the highest comprehensive matching degree as the to-be-determined sequential combination corresponding to the candidate proxy node combination; Perform a comparison process based on the comprehensive matching degrees of the to-be-determined sequential combinations, and determine the to-be-determined sequential combination with the highest comprehensive matching degree as the second proxy node combination.

[0010] Optionally, the fusing the first processing result and the second processing result to obtain the final processing result of the service data includes: Determine a first weight of the first processing result according to the processing duration corresponding to the first processing result, and determine a second weight of the second processing result according to the historical processing success rate associated with the second proxy node combination corresponding to the second processing result; Perform a fusion calculation on the first processing result and the second processing result according to the first weight and the second weight to obtain the final processing result of the service data.

[0011] In a second aspect, an embodiment of the present application further provides a complex service processing system based on a distributed intelligent proxy system, including: An acquisition module, configured to acquire service data to be processed and corresponding service processing requirements; A task parsing module, configured to perform a parsing process on the service processing requirements to obtain a plurality of subtasks; A type allocation module, configured to perform proxy type allocation on each of the subtasks according to the attributes of the service data to obtain corresponding proxy types, where the proxy types include single-step proxy types and sequential proxy types; A proxy node determination module, configured to determine a first distributed proxy node among a plurality of distributed proxy nodes based on a load resource prediction model and the load requirements associated with the subtasks of the single-step proxy type; A task sending module, configured to send the subtasks of the single-step proxy type to the first distributed proxy node for processing to obtain a first processing result; A node combination determination module, configured to screen out multiple candidate agent node combinations for subtasks of the sequential agent type based on the load rate and communication delay time of each of the unselected distributed agent nodes, and determine a second agent node combination based on the matching degree between the candidate agent nodes in each of the candidate agent node combinations; The task sending module is further configured to send the subtasks of the sequential agent type to the second agent node combination for processing to obtain a second processing result; A fusion processing module, configured to perform fusion processing on the first processing result and the second processing result to obtain a final processing result of the service data.

[0012] In a third aspect, an embodiment of the present application further provides a complex service processing device based on a distributed intelligent agent system, and the device includes: One or more processors; A storage device, configured to store one or more programs, When the one or more programs are executed by the one or more processors, the one or more processors implement the complex service processing method based on the distributed intelligent agent system according to the embodiment of the present application.

[0013] In a fourth aspect, an embodiment of the present application further provides a storage medium storing computer-executable instructions, and the computer-executable instructions are used to execute the complex service processing method based on the distributed intelligent agent system according to the embodiment of the present application when executed by a computer processor.

[0014] In the embodiments of the present application, by obtaining the service data to be processed and the corresponding service processing requirements, parsing the service processing requirements to obtain multiple subtasks, allocating an agent type to each subtask according to the attributes of the service data to obtain the corresponding agent type, where the agent type includes a single-step agent type and a sequential agent type, determining a first distributed agent node among multiple distributed agent nodes based on the load resource prediction model and the load requirements associated with the subtasks of the single-step agent type, and sending the subtasks of the single-step agent type to the first distributed agent node for processing to obtain a first processing result, screening out multiple candidate agent node combinations for the subtasks of the sequential agent type according to the load rates and communication delay times of the unselected distributed agent nodes, determining a second agent node combination based on the matching degrees among the candidate agent nodes in each candidate agent node combination, sending the subtasks of the sequential agent type to the second agent node combination for processing to obtain a second processing result, and performing a fusion process on the first processing result and the second processing result to obtain the final processing result of the service data. This solution solves the problem in the prior art of low processing efficiency for complex service problems and inability to meet the processing requirements, and can achieve efficient processing of service data, while reducing the processing cost while ensuring the processing efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 It is a flowchart of a method for processing complex services based on a distributed intelligent agent system provided by an embodiment of the present application; Figure 2 It is a flowchart of a method for processing complex services based on a distributed intelligent agent system that includes determining a first distributed agent node provided by an embodiment of the present application; Figure 3 It is a flowchart of a method for processing complex services based on a distributed intelligent agent system that includes determining multiple candidate agent node combinations provided by an embodiment of the present application; Figure 4 It is a flowchart of a method for processing complex services based on a distributed intelligent agent system that includes determining a second agent node combination provided by an embodiment of the present application; Figure 5 It is a flowchart of a method for processing complex services based on a distributed intelligent agent system that includes determining a final processing result provided by an embodiment of the present application; Figure 6 It is a block diagram of the module structure of a complex service processing system based on a distributed intelligent agent system provided by an embodiment of the present application; Figure 7 It is a schematic diagram of the structure of a complex service processing device based on a distributed intelligent agent system provided by an embodiment of the present application. DETAILED DESCRIPTION

[0016] The following further elaborates on the embodiments of the present application in conjunction with the accompanying drawings and examples. It can be understood that the specific embodiments described herein are only used to explain the embodiments of the present application, rather than limiting the embodiments of the present application. Additionally, it should be noted that for ease of description, only parts related to the embodiments of the present application rather than all structures are shown in the accompanying drawings.

[0017] The terms "first", "second", etc. in the description and claims of the present application are used to distinguish similar objects, rather than to describe a specific order or sequence. It should be understood that such data 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 of the same type, and the number of objects is not limited. For example, the first object can be one or multiple. In addition, "and / or" in the description and claims indicates at least one of the connected objects, and the character " / ", generally represents an "or" relationship between the associated objects before and after.

[0018] A complex business processing method based on a distributed intelligent agent system provided by an embodiment of the present application can be applied to processing scenarios of complex businesses such as high-frequency quantitative trading decision-making chains, multi-modal financial risk control, and supply chain intelligent scheduling. For a complex business processing method based on a distributed intelligent agent system provided by an embodiment of the present application, the execution entity of each step is a server.

[0019] Figure 1 It is a flowchart of a complex business processing method based on a distributed intelligent agent system provided by an embodiment of the present application. As Figure 1 shown, it specifically includes: Step S101: Obtain the business data to be processed and the corresponding business processing requirements, perform parsing processing on the business processing requirements to obtain multiple subtasks, and assign agent types to each subtask according to the attributes of the business data to obtain corresponding agent types, where the agent types include single-step agent types and sequential agent types.

[0020] Among them, business data is used to represent the relevant data of the business to be processed. The business processing requirement can be the goal to be achieved by processing the business data, and this business processing requirement can be a text description. For example, the business processing requirement is to analyze the order data of a certain e-commerce platform today in real time to identify high-value users and push coupons. Multiple subtasks are obtained through parsing and processing using this business processing requirement. A subtask can be each task that needs to be completed to achieve the business processing requirement. The attributes of the business data can be used to assign the proxy type to each of these subtasks. The proxy type refers to the task processing mode divided according to the structure and execution logic of the subtask. This proxy type can include a single-step proxy type and a sequential proxy type. The single-step proxy type refers to the type of subtask that can be completed in a single operation without relying on other steps or intermediate results. The sequential proxy type refers to the type of subtask that requires the cooperation of multiple steps, or the output of the previous step is used as the input of the next step. In one embodiment, after obtaining the business data to be processed and the corresponding business processing requirement, natural language processing is performed on the business processing requirement to obtain the corresponding multiple subtasks, and it is determined whether there is a dependency relationship between the subtasks according to the dependency relationship between the sub-data in the business data. The subtasks with a dependency relationship are determined as the sequential proxy type, and the subtasks without a dependency relationship are determined as the single-step proxy type. Exemplarily, the business processing requirement is to judge financial risks, the business data includes a transaction amount table, a transaction IP address table, a user login information table, a user browsing information table, and a payment information table. There is a dependency relationship between the user login information table and the user browsing information table, and there is a dependency relationship between the user browsing information table and the payment information table. Natural language processing is performed on the business processing requirement to obtain the corresponding multiple subtasks as detecting the transaction amount, detecting the transaction IP address, analyzing the user login behavior, analyzing the user browsing behavior, and analyzing the user payment behavior. Then, detecting the transaction amount and detecting the transaction IP address are subtasks of the single-step proxy type, and analyzing the user login behavior → analyzing the user browsing behavior → analyzing the user payment behavior are subtasks of the sequential proxy type.

[0021] Step S102: Determine the first distributed proxy node among multiple distributed proxy nodes based on the load resource prediction model and the load requirements associated with the subtasks of the single-step proxy type, and send the subtasks of the single-step proxy type to the first distributed proxy node for processing to obtain the first processing result.

[0022] Among them, the load resource prediction model can be a model used to predict the load resources of a distributed agent node within a certain period of time. The load demand refers to the demand of a subtask of a single-step agent type for the load resources of the distributed agent node. Using the load resource prediction model and the load demand, the distributed agent node for processing the subtask of the corresponding single-step agent type can be determined among multiple distributed agent nodes. A distributed agent node refers to an intelligent computing unit deployed in a distributed system with autonomous decision-making capabilities, which realizes complex business processing through a collaborative working mechanism. Each such distributed agent node can operate independently or cooperate with other nodes in the system through a communication protocol to form a unified processing network. In one embodiment, a way to determine the first distributed agent node can be to input the current comprehensive load information and the time period of multiple distributed agent nodes into the load resource prediction model to obtain the load resource prediction information of each distributed agent node, compare the load resource prediction information of each distributed agent node with the load demand associated with the subtask of the single-step agent type, and determine the optimal distributed agent node according to the comparison result and determine it as the first distributed agent node for the subtask of the single-step agent type.

[0023] Step S103: Screen out multiple candidate agent node combinations for the subtasks of the sequential agent type according to the load rates and communication delay times of the unselected distributed agent nodes, and determine the second agent node combination based on the matching degrees between the candidate agent nodes in each candidate agent node combination.

[0024] Among them, the load rate is used to represent the ratio of the actual workload of the distributed agent node for processing requests, data, or computing tasks currently to its theoretical maximum carrying capacity. The communication delay time can be the delay time for the distributed agent node to parse, calculate, or forward data packets. Using the load rates and communication delay times of the unselected distributed agent nodes, multiple candidate agent node combinations for the subtasks of the sequential agent type can be screened out. The candidate agent node combination is used to represent the combination of distributed agent nodes to be selected for processing the subtasks of the sequential agent type. In one embodiment, the subtasks of the sequential agent type include multiple minimum division tasks executed sequentially. A way to screen multiple candidate agent node combinations can be to calculate the corresponding real-time scores according to the load rates and communication delay times of the unselected distributed agent nodes, screen out multiple candidate agent nodes with real-time scores higher than a preset threshold, and perform combination processing on the multiple candidate agent nodes according to the number of minimum division tasks to obtain multiple candidate agent node combinations. In one embodiment, a way to determine the second agent node combination can be to calculate the average value of the matching degrees between the candidate agent nodes in each candidate agent node combination, and determine the candidate agent node combination with the highest average value as the second agent node combination.

[0025] Step S104: Send the subtasks of the sequential proxy type to the second proxy node combination for processing to obtain a second processing result, and fuse the first processing result and the second processing result to obtain the final processing result of the service data.

[0026] Among them, after determining the second proxy node combination, send the subtasks of the sequential proxy type to the second proxy node combination for processing to obtain a second processing result. The second processing result can be fused with the first processing result to obtain the final processing result of the service data. In one embodiment, a fusion processing method may be to determine a first weight according to the processing duration and historical success rate of the first distributed proxy node corresponding to the first processing result, determine a second weight according to the processing duration and historical processing success rate of the second proxy node combination corresponding to the second processing result, and perform a fusion calculation on the first processing result and the second processing result according to the first weight and the second weight to obtain the final processing result of the service data. By assigning corresponding weights to each processing result based on the processing duration and historical processing success rate and performing a fusion calculation to obtain the final processing result of the service data, the rationality and accuracy of the final processing result can be improved.

[0027] As can be seen from the above, by obtaining the service data to be processed and the corresponding service processing requirements, parsing the service processing requirements to obtain multiple subtasks, assigning proxy types to each subtask according to the attributes of the service data to obtain corresponding proxy types, where the proxy types include single-step proxy type and sequential proxy type, determining the first distributed proxy node among multiple distributed proxy nodes based on the load resource prediction model and the load requirements associated with the subtasks of the single-step proxy type, sending the subtasks of the single-step proxy type to the first distributed proxy node for processing to obtain a first processing result, screening out multiple candidate proxy node combinations for the subtasks of the sequential proxy type according to the load rates and communication delay times of the unselected distributed proxy nodes, determining the second proxy node combination based on the matching degree between the candidate proxy nodes in each candidate proxy node combination, sending the subtasks of the sequential proxy type to the second proxy node combination for processing to obtain a second processing result, and fusing the first processing result and the second processing result to obtain the final processing result of the service data. This solution solves the problem of low processing efficiency for complex service problems in the prior art and inability to meet processing requirements, and can achieve efficient processing of service data, while reducing processing costs while ensuring processing efficiency.

[0028] Figure 2 The flowchart of a complex service processing method based on a distributed intelligent proxy system including determining the first distributed proxy node provided by an embodiment of the present application is as Figure 2 shown, and specifically includes: Step S201: Obtain the business data to be processed and the corresponding business processing requirements, parse the business processing requirements to obtain multiple subtasks, and allocate proxy types to each subtask according to the attributes of the business data to obtain the corresponding proxy types, where the proxy types include single-step proxy types and sequential proxy types.

[0029] Step S202: Input the current comprehensive load information and the time period of multiple distributed proxy nodes into the load resource prediction model to obtain the load resource prediction information of each distributed proxy node. Determine the first distributed proxy node of the subtask corresponding to the single-step proxy type according to the load resource prediction information and the load demand of each distributed proxy node, and send the subtask of the single-step proxy type to the first distributed proxy node for processing to obtain the first processing result.

[0030] Among them, the comprehensive load information can be the comprehensive information of each load resource of the distributed proxy node currently, such as load resources like CPU occupancy rate, queue length, memory usage rate, etc. This comprehensive load information can also be the comprehensive load score evaluated from each load resource of the distributed proxy node currently. The time period is used to represent the time period in which multiple distributed proxy nodes are currently located, such as time periods like working day periods, rest day periods, holiday periods, etc. By inputting the current comprehensive load information and the time period of multiple distributed proxy nodes into the load resource prediction model, the load resource prediction information of each distributed proxy node can be obtained. The load resource prediction information is used to represent the load resource information of the predicted distributed proxy node in the next time period. Using this load resource prediction information and the load demand associated with the subtask of the single-step proxy type, the first distributed proxy node can be determined, and sending the subtask of the single-step proxy type to this first distributed proxy node for processing can obtain the first processing result.

[0031] Optionally, a way to determine the first distributed agent node can be as follows: when there are multiple sub-tasks of the single-step agent type, compare the load requirements associated with each sub-task of the single-step agent type, perform a priority ranking based on the comparison result to obtain the allocation priorities of each sub-task of the single-step agent type, compare the load requirements of the sub-task with the lowest allocation priority with the load resource prediction information of each distributed agent node to obtain an optional agent node sequence combination, and determine the first distributed agent node of each sub-task of the single-step agent type according to the allocation priority and the optional agent node sequence combination. Exemplarily, three sub-tasks of the single-step agent type are sub-task 1, sub-task 2, and sub-task 3 respectively. The load requirement of sub-task 1 is that the CPU occupancy rate is less than 30% and the memory usage rate is less than 40%. The load requirement of sub-task 2 is that the CPU occupancy rate is less than 40% and the memory usage rate is less than 50%. The load requirement of sub-task 3 is that the CPU occupancy rate is less than 45% and the memory usage rate is less than 50%. Then the allocation priorities of each sub-task of the single-step agent type are sub-task 1 > sub-task 2 > sub-task 3. The six distributed agent nodes are node a, node b, node c, node d, node e, and node f respectively. The load resource prediction information of node a is that the CPU occupancy rate is 25% and the memory usage rate is 35%. The load resource prediction information of node b is that the CPU occupancy rate is 20% and the memory usage rate is 25%. The load resource prediction information of node c is that the CPU occupancy rate is 35% and the memory usage rate is 45%. The load resource prediction information of node d is that the CPU occupancy rate is 35% and the memory usage rate is 40%. The load resource prediction information of node e is that the CPU occupancy rate is 50% and the memory usage rate is 40%. The load resource prediction information of node d is that the CPU occupancy rate is 50% and the memory usage rate is 55%. The distributed agent nodes that meet the load requirements of sub-task 3 are node a, node b, node c, and node d. Sort the distributed agent nodes that meet the requirements according to the degree of compliance to obtain the optional agent node sequence combination as node b > node a > node d > node c. According to the allocation priority, determine that the first distributed agent node of sub-task 1 is node b, the first distributed agent node of sub-task 2 is node a, and the first distributed agent node of sub-task 3 is node d.

[0032] In another embodiment, a way to determine the first distributed agent node can be to determine the load resource prediction information among the load resource prediction information of each distributed agent node that is closest to the load requirement of the sub-task of the single-step agent type, and determine the closest load resource prediction information as the first distributed agent node of the sub-task of the single-step agent type.

[0033] Step S203: Based on the load rates and communication delay times of each unselected distributed agent node, filter out multiple candidate agent node combinations for the sub-tasks of the sequential agent type, and determine the second agent node combination based on the matching degree among the candidate agent nodes in each candidate agent node combination.

[0034] Step S204: Send the sub-tasks of the sequential agent type to the second agent node combination for processing to obtain a second processing result, and fuse the first processing result and the second processing result to obtain the final processing result of the service data.

[0035] As can be seen from the above, by inputting the current comprehensive load information and the time period of each distributed agent node into the load resource prediction model, the load resource prediction information of each distributed agent node is obtained. According to the load resource prediction information and load requirements of each distributed agent node, the first distributed agent node for the sub-tasks of the single-step agent type is determined, and the sub-tasks of the single-step agent type are sent to the first distributed agent node for processing to obtain a first processing result. This solution can ensure load balancing and improve task processing efficiency by predicting the load resources of each allocated agent node and determining the corresponding distributed agent node according to the load requirements of the single-step agent type sub-tasks.

[0036] Figure 3 The figure is a flowchart of a complex service processing method based on a distributed intelligent agent system provided by an embodiment of the present application, which includes determining multiple candidate agent node combinations. As Figure 3 shown, it specifically includes: Step S301: Obtain the service data to be processed and the corresponding service processing requirements, perform parsing processing on the service processing requirements to obtain multiple sub-tasks, and perform agent type allocation on each sub-task according to the attributes of the service data to obtain the corresponding agent type, where the agent type includes a single-step agent type and a sequential agent type.

[0037] Step S302: Determine the first distributed agent node among multiple distributed agent nodes based on the load resource prediction model and the load requirements associated with the sub-tasks of the single-step agent type, and send the sub-tasks of the single-step agent type to the first distributed agent node for processing to obtain a first processing result.

[0038] Step S303: Filter out multiple candidate agent nodes with a load rate lower than the first threshold and a communication delay time lower than the second threshold from the unselected distributed agent nodes, perform combination processing on the multiple candidate agent nodes according to the number of minimum divided tasks to obtain multiple candidate agent node combinations, and determine the second agent node combination based on the matching degree among the candidate agent nodes in each candidate agent node combination.

[0039] Among them, the subtasks of the sequential proxy type include multiple minimally divided tasks that are executed sequentially. The first threshold is a threshold set according to actual requirements for comparing the load ratio. The second threshold is a threshold set according to actual requirements for comparing the communication delay time. By using the first threshold and the second threshold to compare with each unselected distributed proxy node, a load ratio and a communication delay time can be obtained for comparison processing to obtain multiple candidate proxy nodes. By combining the multiple candidate proxy nodes according to the number of minimally divided tasks in the subtasks of the sequential proxy type, multiple candidate proxy node combinations can be obtained. Exemplarily, each unselected distributed proxy node is respectively node 1, node 2, node 3, node 4, and node 5, and the minimally divided tasks in the subtasks of the sequential proxy type are task a → task b → task c in sequence. The candidate proxy nodes with a load ratio lower than the first threshold and a communication delay time lower than the second threshold are selected as node 1, node 2, node 3, and node 4. According to the number of minimally divided tasks, multiple candidate proxy node combinations are obtained by combining the multiple candidate proxy nodes, which are respectively node 1, 2, 3, node 1, 2, 4, node 1, 3, 4, and node 2, 3, 4.

[0040] Step S304: Send the subtasks of the sequential proxy type to the second proxy node combination for processing to obtain a second processing result, and fuse the first processing result and the second processing result to obtain the final processing result of the service data.

[0041] As can be seen from the above, by screening out multiple candidate proxy nodes with a load ratio lower than the first threshold and a communication delay time lower than the second threshold according to each unselected distributed proxy node, and combining the multiple candidate proxy nodes according to the number of minimally divided tasks, multiple candidate proxy node combinations are obtained. This solution determines the candidate distributed proxy node combination for processing the subtasks of the sequential proxy type through the load ratio and the communication delay time, which can ensure the rationality of the selected distributed proxy node combination and improve the processing efficiency.

[0042] Figure 4 It is a flowchart of a complex service processing method based on a distributed intelligent proxy system including determining a second proxy node combination provided by an embodiment of the present application. As Figure 4 shown, it specifically includes: Step S401: Obtain the service data to be processed and the corresponding service processing requirements, perform parsing processing on the service processing requirements to obtain multiple subtasks, and perform proxy type allocation on each subtask according to the attributes of the service data to obtain the corresponding proxy type. Among them, the proxy type includes a single-step proxy type and a sequential proxy type.

[0043] Step S402: Determine the first distributed agent node among multiple distributed agent nodes based on the load resource prediction model and the load requirements associated with the subtasks of the single-step agent type, and send the subtasks of the single-step agent type to the first distributed agent node for processing to obtain the first processing result.

[0044] Step S403: Screen out multiple candidate agent node combinations for the subtasks of the sequential agent type according to the load rates and communication delay times of the unselected distributed agent nodes, arrange the candidate agent nodes in each of the candidate agent node combinations in sequence to obtain the corresponding multiple sequential agent node sequences, and determine the second agent node combination according to the matching degrees between the candidate agent nodes and the multiple sequential agent node sequences of each candidate agent node combination.

[0045] Among them, the matching degree can be the compatibility degree between distributed agent nodes. Optionally, a way to determine the second agent node combination can be to superimpose the matching degrees between the sequentially connected candidate agent nodes in the multiple sequential agent node sequences of the candidate agent node combination to obtain the comprehensive matching degree, determine the sequential combination to be determined corresponding to the candidate agent node combination with the highest comprehensive matching degree, compare and process according to the comprehensive matching degrees of each sequential combination to be determined, and determine the sequential combination to be determined with the highest comprehensive matching degree as the second agent node combination. Exemplarily, for the candidate agent node combination A which is distributed agent nodes 1, 2, and 3, the sequential agent node sequence a is 1→2→3, the sequential agent node sequence b is 1→3→2, the sequential agent node sequence c is 2→1→3, the sequential agent node sequence d is 2→3→1, the sequential agent node sequence e is 3→1→2, the sequential agent node sequence f is 3→2→1. The matching degree between distributed agent node 1 and distributed agent node 2 in the sequential agent node sequence a is 90, and the matching degree between distributed node 2 and distributed node 3 is 95, so the comprehensive matching degree is 185. Calculate the comprehensive matching degrees of the sequential agent node sequences b, c, d, e, and f in the same way, which are 190, 180, 175, 196, and 185 respectively. The sequential agent node sequence with the highest comprehensive matching degree is 3→1→2, and 3→1→2 is determined as the sequential combination to be determined for the candidate agent node combination A. In the same way, the sequential combinations to be determined for the candidate agent node combinations B, C, and D are 1→2→4, 1→3→4, and 2→3→4 respectively, and the corresponding comprehensive matching degrees are 193, 190, and 188 respectively. Since the comprehensive matching degree of the sequential combination to be determined 3→1→2 is the highest, 3→1→2 is determined as the second agent node combination.

[0046] In another embodiment, a method for determining a second combination of proxy nodes may be as follows: calculate the average matching degree of the matching degrees between the candidate proxy nodes connected in sequence in multiple sequential proxy node sequences of the candidate proxy node combinations, determine the to-be-determined sequential combination corresponding to the candidate proxy node combination as the sequential proxy node sequence with the highest average matching degree, perform a comparison process based on the average matching degrees of the to-be-determined sequential combinations, and determine the to-be-determined sequential combination with the highest average matching degree as the second combination of proxy nodes.

[0047] Step S404: Send the sub-tasks of the sequential proxy type to the second combination of proxy nodes for processing to obtain a second processing result, and fuse the first processing result and the second processing result to obtain the final processing result of the service data.

[0048] As can be seen from the above, by arranging the candidate proxy nodes in each candidate proxy node combination in sequence respectively to obtain the corresponding multiple sequential proxy node sequences, the second combination of proxy nodes is determined according to the matching degrees between the candidate proxy nodes and the multiple sequential proxy node sequences of each candidate proxy node combination. This solution can improve the rationality of the selected distributed proxy node combination and the task processing efficiency by determining the distributed proxy node combination for processing the sequential proxy type sub-tasks based on the matching degrees between the candidate proxy nodes.

[0049] Figure 5 The figure is a flowchart of a complex service processing method based on a distributed intelligent proxy system provided by an embodiment of the present application, which includes determining a final processing result. As Figure 5 shown, it specifically includes: Step S501: Obtain the service data to be processed and the corresponding service processing requirements, perform parsing processing on the service processing requirements to obtain multiple sub-tasks, and perform proxy type allocation on each sub-task according to the attributes of the service data to obtain the corresponding proxy types, where the proxy types include single-step proxy types and sequential proxy types.

[0050] Step S502: Determine a first distributed proxy node among multiple distributed proxy nodes based on the load resource prediction model and the load requirements associated with the sub-tasks of the single-step proxy type, and send the sub-tasks of the single-step proxy type to the first distributed proxy node for processing to obtain a first processing result.

[0051] Step S503: Screen out multiple candidate proxy node combinations for the sub-tasks of the sequential proxy type according to the load rates and communication delay times of the unselected distributed proxy nodes, and determine the second combination of proxy nodes based on the matching degrees between the candidate proxy nodes in each candidate proxy node combination.

[0052] Step S504: Send the sub-tasks of the sequential proxy type to the second proxy node combination for processing to obtain a second processing result. Determine the first weight of the first processing result according to the processing duration corresponding to the first processing result. Determine the second weight of the second processing result according to the historical processing success rate associated with the second proxy node combination corresponding to the second processing result. Perform a fusion calculation on the first processing result and the second processing result according to the first weight and the second weight to obtain the final processing result of the service data.

[0053] Among them, the processing duration is used to represent the duration taken by the first distributed proxy node to process the sub-tasks of the corresponding single-step proxy type to obtain the first processing result. The first weight of the first processing result can be determined using this processing duration. The historical processing success rate can be the success rate when the second proxy node combination processes historical tasks. The second weight of the second processing result can be determined according to this historical processing success rate. The final processing result of the service data can be obtained through a fusion calculation using this first weight and this second weight. Exemplarily, the sub-task of the single-step proxy type is to detect the risk of transaction amount, and the sub-task of the sequential proxy type is to analyze the risk of user behavior (analyze user login behavior → analyze user browsing behavior → analyze user payment behavior). The first processing result is a transaction risk score of 9.0, the second processing result is a behavior risk score of 8.0, the processing duration of the first result is 2s, and the corresponding weight is 0.4. The historical processing success rate associated with the second proxy node combination corresponding to the second processing result is 80%, and the corresponding weight is 0.6. Then the final risk score is 9.0 * 0.4 + 8.0 * 0.6 = 0.84.

[0054] As can be seen from the above, by determining the first weight of the first processing result according to the processing duration corresponding to the first processing result, determining the second weight of the second processing result according to the historical processing success rate associated with the second proxy node combination corresponding to the second processing result, and performing a fusion calculation on the first processing result and the second processing result according to the first weight and the second weight to obtain the final processing result of the service data. This solution assigns corresponding weights to each processing result through the processing duration and the historical processing success rate, and performs a fusion calculation to obtain the final processing result of the service data, which can improve the rationality and accuracy of the final processing result.

[0055] Figure 6 It is a block diagram of the module structure of a complex service processing system based on a distributed intelligent proxy system provided by an embodiment of the present application. This system is used to execute a complex service processing method based on a distributed intelligent proxy system provided by the above embodiment, and has the corresponding functional modules and beneficial effects for executing the method. As Figure 6 shown, this system specifically includes: An acquisition module 101, configured to acquire the service data to be processed and the corresponding service processing requirements; The task parsing module 102 is configured to parse and process the service processing requirements to obtain multiple subtasks; The type allocation module 103 is configured to allocate an agent type to each of the subtasks according to the attributes of the service data to obtain a corresponding agent type, where the agent type includes a single-step agent type and a sequential agent type; The agent node determination module 104 is configured to determine a first distributed agent node among multiple distributed agent nodes based on a load resource prediction model and the load requirements associated with the subtasks of the single-step agent type; The task sending module 105 is configured to send the subtasks of the single-step agent type to the first distributed agent node for processing to obtain a first processing result; The node combination determination module 106 is configured to screen out multiple candidate agent node combinations for the subtasks of the sequential agent type according to the load rates and communication delay times of the unselected distributed agent nodes, and determine a second agent node combination based on the matching degrees among the candidate agent nodes in each of the candidate agent node combinations; The task sending module 105 is further configured to send the subtasks of the sequential agent type to the second agent node combination for processing to obtain a second processing result; The fusion processing module 107 is configured to perform fusion processing on the first processing result and the second processing result to obtain a final processing result of the service data.

[0056] As can be seen from the above solution, by obtaining the service data to be processed and the corresponding service processing requirements, parsing and processing the service processing requirements to obtain multiple subtasks, allocating an agent type to each subtask according to the attributes of the service data to obtain a corresponding agent type, where the agent type includes a single-step agent type and a sequential agent type, determining a first distributed agent node among multiple distributed agent nodes based on a load resource prediction model and the load requirements associated with the subtasks of the single-step agent type, sending the subtasks of the single-step agent type to the first distributed agent node for processing to obtain a first processing result, screening out multiple candidate agent node combinations for the subtasks of the sequential agent type according to the load rates and communication delay times of the unselected distributed agent nodes, determining a second agent node combination based on the matching degrees among the candidate agent nodes in each of the candidate agent node combinations, sending the subtasks of the sequential agent type to the second agent node combination for processing to obtain a second processing result, and performing fusion processing on the first processing result and the second processing result to obtain a final processing result of the service data. This solution solves the problem of low processing efficiency for complex service problems in the prior art and inability to meet the processing requirements, and can achieve efficient processing of service data, while reducing the processing cost while ensuring the processing efficiency.

[0057] In a possible embodiment, the proxy node determination module 104 is specifically configured to: Input the current comprehensive load information and the time period of multiple said distributed proxy nodes into the load resource prediction model to obtain the load resource prediction information of each said distributed proxy node; According to the load resource prediction information of each said distributed proxy node and the load demand, determine the first distributed proxy node for the sub-task of the corresponding single-step proxy type.

[0058] In a possible embodiment, the proxy node determination module 104 is further configured to: In the case where there are multiple sub-tasks of the single-step proxy type, compare and process the load demands associated with each said sub-task of the single-step proxy type, and perform priority sorting according to the comparison and processing results to obtain the allocation priorities of each said sub-task of the single-step proxy type; Compare the load demand of the sub-task with the lowest allocation priority with the load resource prediction information of each said distributed proxy node to obtain an optional proxy node sequence combination, and determine the first distributed proxy node for each said sub-task of the single-step proxy type according to the allocation priority and the optional proxy node sequence combination.

[0059] In a possible embodiment, the node combination determination module 106 is specifically configured to: Filter out multiple candidate proxy nodes with a load rate lower than a first threshold and a communication delay time lower than a second threshold from each said unselected distributed proxy node, and perform combination processing on the multiple said candidate proxy nodes according to the number of the smallest divided tasks to obtain multiple candidate proxy node combinations.

[0060] In a possible embodiment, the node combination determination module 106 is further configured to: Arrange each said candidate proxy node in each said candidate proxy node combination in sequence to obtain corresponding multiple sequential proxy node sequences; Determine the second proxy node combination according to the matching degree between each said candidate proxy node and the multiple said sequential proxy node sequences of each said candidate proxy node combination.

[0061] In a possible embodiment, the node combination determination module 106 is further configured to: Superimpose the matching degrees between the sequentially connected candidate proxy nodes in the multiple sequential proxy node sequences of the candidate proxy node combination to obtain a comprehensive matching degree, and determine the sequential proxy node sequence with the highest comprehensive matching degree as the to-be-determined sequential combination of the corresponding candidate proxy node combination; Perform comparison processing based on the comprehensive matching degrees of each combination to be determined in order, and determine the combination to be determined in order with the highest comprehensive matching degree as the second proxy node combination.

[0062] In a possible embodiment, the fusion processing module 107 is specifically configured to: Determine a first weight of the first processing result according to the processing duration corresponding to the first processing result, and determine a second weight of the second processing result according to the historical processing success rate associated with the second proxy node combination corresponding to the second processing result; Perform fusion calculation on the first processing result and the second processing result according to the first weight and the second weight to obtain the final processing result of the service data.

[0063] Figure 7 The figure is a schematic structural diagram of a complex service processing device based on a distributed intelligent agent system provided by an embodiment of the present application. As Figure 7 shown, the device includes a processor 201, a memory 202, an input device 203, and an output device 204; the number of processors 201 in the device may be one or more. Figure 7 Taking one processor 201 as an example; the processor 201, the memory 202, the input device 203, and the output device 204 in the device may be connected through a bus or other means. Figure 7 Taking connection through a bus as an example. The memory 202, as a computer-readable storage medium, can be used to store software programs, computer-executable programs, and modules, such as program instructions or modules corresponding to a complex service processing method based on a distributed intelligent agent system in an embodiment of the present application. The processor 201 executes various functional applications and data processing of the device by running the software programs, instructions, and modules stored in the memory 202, that is, implements the above-mentioned complex service processing method based on a distributed intelligent agent system. The input device 203 can be used to receive input digital or character information, and generate key signal inputs related to user settings and function controls of the device. The output device 204 may include a display device such as a display screen.

[0064] An embodiment of the present application further provides a storage medium containing computer-executable instructions, and the computer-executable instructions are used to execute a complex service processing method based on a distributed intelligent agent system when executed by a computer processor. The method includes: Obtain service data to be processed and corresponding service processing requirements, perform parsing processing on the service processing requirements to obtain a plurality of subtasks, and perform proxy type allocation on each subtask according to the attributes of the service data to obtain corresponding proxy types, where the proxy types include single-step proxy types and sequential proxy types; Determine a first distributed agent node among multiple distributed agent nodes based on the load resource prediction model and the load requirements associated with the subtasks of the single-step agent type, and send the subtasks of the single-step agent type to the first distributed agent node for processing to obtain a first processing result; Filter out multiple candidate agent node combinations for the subtasks of the sequential agent type according to the load rates and communication delay times of the unselected distributed agent nodes, and determine a second agent node combination based on the matching degrees among the candidate agent nodes in each of the candidate agent node combinations; Send the subtasks of the sequential agent type to the second agent node combination for processing to obtain a second processing result, and perform a fusion process on the first processing result and the second processing result to obtain the final processing result of the service data.

[0065] It should be noted that in the embodiments of the complex service processing method system based on the distributed intelligent agent system described above, the various units and modules included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be realized; in addition, the specific names of the functional units are only for the convenience of mutual distinction and do not limit the protection scope of the embodiments of the present application.

[0066] Note that the above is only a preferred embodiment of the embodiments of the present application and the applied technical principles. Those skilled in the art will understand that the embodiments of the present application are not limited to the specific embodiments described herein, and various obvious changes, re-adjustments and substitutions can be made by those skilled in the art without departing from the protection scope of the embodiments of the present application. Therefore, although the embodiments of the present application have been described in more detail through the above embodiments, the embodiments of the present application are not limited to the above embodiments. Without departing from the concept of the embodiments of the present application, more other equivalent embodiments can be included, and the scope of the embodiments of the present application is determined by the scope of the appended claims.

Claims

1. A complex service processing method based on a distributed intelligent agent system, characterized in that, The method includes: Obtaining service data to be processed and corresponding service processing requirements, parsing the service processing requirements to obtain multiple subtasks, and allocating proxy types to each of the subtasks according to the attributes of the service data to obtain corresponding proxy types, where the proxy types include single-step proxy types and sequential proxy types; Determining a first distributed proxy node among multiple distributed proxy nodes based on a load resource prediction model and the load requirements associated with the subtasks of the single-step proxy type, and sending the subtasks of the single-step proxy type to the first distributed proxy node for processing to obtain a first processing result; Filtering out multiple candidate proxy node combinations for the subtasks of the sequential proxy type according to the load rates and communication delay times of the unselected distributed proxy nodes, and determining a second proxy node combination based on the matching degrees among the candidate proxy nodes in each of the candidate proxy node combinations; Sending the subtasks of the sequential proxy type to the second proxy node combination for processing to obtain a second processing result, and fusing the first processing result and the second processing result to obtain the final processing result of the service data.

2. The complex service processing method based on a distributed intelligent agent system according to claim 1, wherein, The determining a first distributed proxy node among multiple distributed proxy nodes based on a load resource prediction model and the load requirements associated with the subtasks of the single-step proxy type includes: Inputting the current comprehensive load information and the time period of the multiple distributed proxy nodes into the load resource prediction model to obtain the load resource prediction information of each of the distributed proxy nodes; Determining the first distributed proxy node for the subtasks of the corresponding single-step proxy type according to the load resource prediction information of each of the distributed proxy nodes and the load requirements.

3. The complex service processing method based on a distributed intelligent agent system according to claim 2, characterized in that The determining the first distributed proxy node for the subtasks of the corresponding single-step proxy type according to the load resource prediction information of each of the distributed proxy nodes and the load requirements includes: In the case where there are multiple subtasks of the single-step proxy type, comparing the load requirements associated with each of the subtasks of the single-step proxy type, and performing priority sorting according to the comparison result to obtain the allocation priorities of each of the subtasks of the single-step proxy type; Comparing the load requirements of the subtask with the lowest allocation priority with the load resource prediction information of each of the distributed proxy nodes to obtain an optional proxy node sequence combination, and determining the first distributed proxy node for each of the subtasks of the single-step proxy type according to the allocation priority and the optional proxy node sequence combination.

4. The complex service processing method based on a distributed intelligent agent system according to any one of claims 1-3, characterized in that, The subtasks of the sequential proxy type include multiple minimum division tasks that are sequentially executed. The filtering out multiple candidate proxy node combinations for the subtasks of the sequential proxy type according to the load rates and communication delay times of the unselected distributed proxy nodes includes: Filter out multiple candidate proxy nodes with a load rate lower than a first threshold and a communication delay time lower than a second threshold according to each of the unselected distributed proxy nodes, and perform combination processing on the multiple candidate proxy nodes according to the number of tasks with the smallest division to obtain multiple combinations of candidate proxy nodes.

5. The complex service processing method based on a distributed intelligent agent system according to any one of claims 1-3, characterized in that The determining the second proxy node combination based on the matching degree between each of the candidate proxy nodes in each of the candidate proxy node combinations includes: Sequentially arrange each of the candidate proxy nodes in each of the candidate proxy node combinations to obtain corresponding multiple ordered proxy node sequences; Determine the second proxy node combination according to the matching degree between each of the candidate proxy nodes and the multiple ordered proxy node sequences of each of the candidate proxy node combinations.

6. The complex service processing method based on a distributed intelligent agent system according to claim 5, wherein The determining the second proxy node combination according to the matching degree between each of the candidate proxy nodes and the multiple ordered proxy node sequences of each of the candidate proxy node combinations includes: Superimpose the matching degrees between each of the candidate proxy nodes connected in sequence in the multiple ordered proxy node sequences of the candidate proxy node combination to obtain a comprehensive matching degree, and determine the ordered proxy node sequence with the highest comprehensive matching degree as the to-be-determined ordered combination corresponding to the candidate proxy node combination; Perform comparison processing according to the comprehensive matching degrees of each of the to-be-determined ordered combinations, and determine the to-be-determined ordered combination with the highest comprehensive matching degree as the second proxy node combination.

7. The complex service processing method based on a distributed intelligent agent system according to any one of claims 1-3, characterized in that The fusing the first processing result and the second processing result to obtain the final processing result of the service data includes: Determine a first weight of the first processing result according to the processing duration corresponding to the first processing result, and determine a second weight of the second processing result according to the historical processing success rate associated with the second proxy node combination corresponding to the second processing result; Perform fusion calculation on the first processing result and the second processing result according to the first weight and the second weight to obtain the final processing result of the service data.

8. A complex business processing system based on a distributed intelligent agent system, characterized in that, Includes: An acquisition module, configured to acquire service data to be processed and corresponding service processing requirements; A task parsing module, configured to perform parsing processing on the service processing requirements to obtain multiple subtasks; A type allocation module, configured to perform proxy type allocation on each of the subtasks according to the attributes of the service data to obtain corresponding proxy types, where the proxy types include single-step proxy types and sequential proxy types; A proxy node determination module, configured to determine a first distributed proxy node among multiple distributed proxy nodes based on a load resource prediction model and the load requirements associated with the subtasks of the single-step proxy type; A task sending module, configured to send the subtasks of the single-step proxy type to the first distributed proxy node for processing to obtain a first processing result; A node combination determination module, configured to screen out multiple candidate proxy node combinations for the subtasks of the sequential proxy type according to the load rate and communication delay time of each of the unselected distributed proxy nodes, and determine a second proxy node combination based on the matching degree between each of the candidate proxy nodes in each of the candidate proxy node combinations; The task sending module is further configured to send the subtasks of the sequential proxy type to the second proxy node combination for processing to obtain a second processing result; The fusion processing module is configured to perform fusion processing on the first processing result and the second processing result to obtain a final processing result of the service data.

9. A complex service processing device based on a distributed intelligent agent system, characterized in that, The device includes: one or more processors; a storage device for storing one or more programs, which when executed by the one or more processors cause the one or more processors to implement the complex service processing method based on a distributed intelligent proxy system according to any one of claims 1-7.

10. A storage medium storing computer-executable instructions, characterized in that, The computer-executable instructions are used to execute the complex service processing method based on a distributed intelligent proxy system according to any one of claims 1-7 when executed by a computer processor.

Images