Automatic robot process scheduling method based on queue processing efficiency

Abstract

The invention relates to a robot process automatic scheduling method based on queue processing efficiency. The method comprises the following steps: step 1, collecting data to be processed and storing the data to be processed in a queue; 2, checking data stored in the queue through an RPA task scheduling engine, and obtaining data needing to be processed in the queue according to the state and priority of the data in the queue; step 3, dynamically adjusting the operation quantity of starting the RPA robot, and calling automatic flow processing data by the RPA robot; and step 4, after the RPA robot executes the task, generating a data processing result and storing the data processing result in the queue. By dynamically adjusting the number of RPA robot operations, the efficiency of mass data processing is improved, and meanwhile the utilization rate of the RPA robot is improved.

Description

Technical Field

[0001] The present invention relates to the field of robotic process automation, and in particular to a robotic process automation scheduling method based on queue processing efficiency. Background Art

[0002] Robotic Process Automation (RPA) is software that runs on personal computers or servers. Its main function is to simulate human control of various system software and automatically execute business processes. It is suitable for large-scale, highly repetitive business scenarios.

[0003] Currently, when RPA processes massive business data, it stores the data to be processed in a queue and processes the data in the queue through the RPA task scheduling engine. During the processing, the job is often triggered after the number of data records in the queue reaches a certain threshold. In order to improve data processing efficiency, existing methods often dynamically adjust task processing by adjusting the threshold of resource allocation, and dynamically adjust and allocate task resources. Existing technologies such as the cluster resource elastic scheduling method disclosed in the invention patent with publication number CN117632414A dynamically adjust the threshold of each task queue allocation to improve data processing efficiency. Summary of the invention

[0004] In order to improve data processing efficiency, the present invention provides a robotic process automation scheduling method based on queue processing efficiency, which avoids backlog of data records in the queue by adjusting the number of jobs in the queue tasks.

[0005] The technical objectives of the present invention are achieved through the following technical solutions:

[0006] A robotic process automation scheduling method based on queue processing efficiency, the method comprising:

[0007] Step 1: Collect the data to be processed and store it in a queue;

[0008] Step 2: Check the data stored in the queue through the RPA task scheduling engine, and obtain the data to be processed in the queue based on the status and priority of the data in the queue;

[0009] Step 3: Dynamically adjust the number of RPA robot jobs started, and the RPA robot calls the automated process to process data;

[0010] Step 4: After the RPA robot completes the task, it generates data processing results and stores them in the queue.

[0011] Furthermore, the amount of data in the queue is N, the data volume threshold for triggering the first RPA robot job is X, the data volume threshold for triggering a new RPA robot job is Z, and the maximum number of RPA robot jobs allowed to run simultaneously is Y by default;

[0012] When ((NX) / Z)>Y, increase the maximum number of RPA robot jobs allowed to run simultaneously.

[0013] Furthermore, the number of RPA robots currently waiting to run is W, and a comprehensive evaluation index is created. The data volume threshold Z for triggering new RPA robot jobs is automatically set according to the current data volume waiting to be processed and the comprehensive evaluation index; when W reaches the set threshold, the Z value is reduced, and when the comprehensive evaluation index is less than the set threshold, the Z value is increased;

[0014] The comprehensive evaluation index is calculated based on the weighted combination of the RPA robot job execution efficiency, resource utilization of the resource pool, and RPA robot job response time, with the sum of the weights being 1.

[0015] Furthermore, the weight of RPA robot job execution efficiency is higher than resource utilization and RPA robot job response time.

[0016] Furthermore, the comprehensive evaluation index = 0.4×RPA robot job execution efficiency + 0.3×resource utilization + 0.3×RPA robot job response time.

[0017] Furthermore, when a new RPA robot job is triggered, the current status of the resource pool and the comprehensive evaluation indicators are monitored to dynamically adjust the number of RPA robot jobs running simultaneously, which can adapt to the ever-changing workload and resource requirements, effectively improve resource utilization and reduce queue response time.

[0018] When resource utilization is lower than the set threshold, increase the number of RPA robot jobs running simultaneously;

[0019] When the RPA robot job response time exceeds the set threshold, increase the number of RPA robot jobs running simultaneously;

[0020] When the resource pool is close to saturation or the comprehensive evaluation index reaches the set threshold, reduce the number of RPA robots running simultaneously.

[0021] Furthermore, when storing the data to be processed in the queue, the data is added to the queue step by step or in batches.

[0022] Furthermore, each piece of data stored in the queue has an identification field, a data priority, a data status and a data processing result.

[0023] Furthermore, the data status includes at least one of pending, being processed, processing exception, and business exception.

[0024] Furthermore, the state of the data in the queue is changed according to the completion result of the RPA robot job execution.

[0025] Compared with the prior art, the present invention has the following beneficial effects:

[0026] The present invention improves the efficiency of massive data processing and the utilization rate of RPA robots by dynamically adjusting the number of RPA robot operations; it more comprehensively evaluates the efficiency of the resource pool through comprehensive evaluation indicators, reduces manual intervention, and improves the automation and efficiency of RPA robot operation scheduling; and it ensures the stability and reliability of the RPA robot operation system by monitoring the resource pool and comprehensive evaluation indicators. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 It is a flow chart of the robotic process automation scheduling method based on queue processing efficiency in the present invention.

[0028] Figure 2 It is a schematic diagram of the process of dynamically adjusting the number of RPA robot operations started in the present invention. DETAILED DESCRIPTION

[0029] The technical solution of the present invention is further described below in conjunction with specific embodiments:

[0030] A robotic process automation scheduling method based on queue processing efficiency, such as Figure 1 As shown, the method includes:

[0031] Step 1: Collect the data to be processed and store it in a queue. The data can be added to the queue in batches or gradually.

[0032] Each piece of data stored in the queue has an identification field, data priority, data status, and data processing result. The data status includes at least one of pending, processing, processing exception, and business exception. The status of the data in the queue changes according to the result of the RPA robot job execution.

[0033] Step 2: Check the data stored in the queue through the RPA task scheduling engine, and obtain the data to be processed in the queue based on the status and priority of the data in the queue;

[0034] Step 3: Dynamically adjust the number of RPA robot jobs started, such as Figure 2As shown, the RPA robot calls the automated process to process data; the automated process includes data processing, system operation, document generation, etc.

[0035] The amount of data in the queue is N, the data volume threshold that triggers the first RPA robot job is X, the data volume threshold that triggers a new RPA robot job is Z, and the maximum number of RPA robot jobs allowed to run simultaneously is Y by default;

[0036] When ((NX) / Z)>Y, increase the maximum number of RPA robot jobs allowed to run simultaneously.

[0037] As a preferred method, it is also possible to consider adjusting the threshold Z for triggering the number of new RPA robot jobs according to the number W of RPA robots currently waiting to run and the creation of a comprehensive evaluation index, wherein the comprehensive evaluation index is calculated according to a weighted combination of the RPA robot job execution efficiency, the resource utilization of the resource pool, and the RPA robot job response time, and the sum of the weights is 1. When W reaches the set threshold, the Z value is reduced, and when the comprehensive evaluation index is less than the set threshold, the Z value is increased. When W reaches the set threshold, it means that the resource utilization is not high at this time. Reducing the Z value can speed up the triggering of the number of RPA robots running at the same time and improve processing efficiency; when the comprehensive evaluation index is less than the set threshold, it means that the current operation of the resource pool is relatively saturated. Increasing the Z value can slow down the number of new RPA robot jobs triggered and reduce the excessive occupation of resources when the RPA robot is working.

[0038] Specifically, the comprehensive evaluation index = RPA robot job execution efficiency × threshold 1 + resource utilization × threshold 2 + RPA robot job response time × threshold 3, threshold 1 + threshold 2 + threshold 3 = 1.

[0039] Among them, the RPA robot job execution efficiency is the time and resource utilization required to complete each job, which can be calculated by the ratio between job completion time and resource consumption.

[0040] Resource utilization refers to the resource utilization in the RPA robot resource pool, including the utilization of CPU, memory, network and other resources. The higher the resource utilization, the more fully the resources are utilized.

[0041] The RPA robot job response time is the time between when the RPA robot job is triggered and when it is actually executed. The shorter the response time, the higher the job efficiency.

[0042] Since the RPA robot job execution efficiency is one of the core indicators of the robot resource pool efficiency, it reflects the effective utilization of resources and the time required to complete the job. Therefore, its weight is set higher than the weight of resource utilization and the weight of RPA robot job response time.

[0043] In one embodiment, weight 1 is set to 0.4, weight 2 to 0.3, and weight 3 to 0.3, and the result is: comprehensive evaluation index = 0.4×RPA robot job execution efficiency + 0.3×resource utilization + 0.3×RPA robot job response time.

[0044] Step 4: After the RPA robot completes the task, it generates data processing results and stores them in the queue.

[0045] As a preferred method, when a new RPA robot job is triggered, the current status of the resource pool and the comprehensive evaluation indicators are monitored to dynamically adjust the number of RPA robot jobs running simultaneously;

[0046] When resource utilization is lower than the set threshold, increase the number of RPA robot jobs running simultaneously;

[0047] When the RPA robot job response time exceeds the set threshold, increase the number of RPA robot jobs running simultaneously;

[0048] When the resource pool is close to saturation or the comprehensive evaluation index reaches the set threshold, the number of RPA robots running simultaneously is reduced. It should be noted that increasing or decreasing the number of RPA robots running simultaneously is adjusted within the range of the maximum number of RPA robot jobs Y allowed to run simultaneously.

[0049] In order to more clearly illustrate the solution of the present application, this embodiment provides 5 examples:

[0050] (1) 60 data items are gradually added to the queue at a rate of 3 data items per second. The amount of data that triggers the first RPA robot job is 31. The amount of data that triggers a new RPA robot job is 10, which means that a new RPA robot job is triggered every time 10 data items are added. The maximum number of RPA robot jobs that can run simultaneously is 3 by default.

[0051] That is, when the amount of data in the queue reaches 31, the first RPA robot job is started. At the 11th second, the amount of data in the queue reaches 33. The first RPA robot starts the job and processes one data in the queue per second. There are 60-31=29 data left, 29 / 10=2.9<3. The default maximum number of RPA robot jobs allowed to run simultaneously is not changed. The number of RPA robots waiting to run is currently 2.

[0052] At the 14th second, the number of data in the queue reaches 42, and the second RPA robot runs the job. Currently, there is one RPA robot waiting to run. At the 17th second, the number of data in the queue reaches 51, and the third RPA robot runs the job. At this time, there is no RPA robot waiting to run.

[0053] The data addition was completed at the 20th second, and there were 42 pieces of data left. The three RPA robots needed another 14 seconds to complete the processing, for a total of 34 seconds.

[0054] The Y value does not change. After the first RPA robot runs a job, if it detects that the current resource utilization is lower than the set threshold, the second and third RPA robots are started in advance to run the job.

[0055] After the first RPA robot runs a job, if the RPA robot job response time is detected to exceed the set threshold, the number of RPA robot jobs running simultaneously is increased;

[0056] After all three RPA robots are running jobs, when it is monitored that the resource pool is close to saturation or the comprehensive evaluation index reaches the set threshold, the number of RPA robots running simultaneously is reduced, so that the third RPA robot and / or the second RPA robot stops working, thereby reducing competitive resource occupation and performance degradation.

[0057] (2) 60 data are added to the queue at a time, and the amount of data that triggers the RPA robot job is 31; the amount of data that triggers a new RPA robot job is 10, that is, every 10 data added will trigger a new RPA robot job; the default maximum number of RPA robot jobs allowed to run simultaneously is 3; when 60 data are added at a time, the three RPA robots all start the job. At this time, there is no RPA robot currently waiting to run. If each RPA robot processes one per second, it will take 20 seconds to complete the processing.

[0058] The Y value does not change. When it is monitored that the resource pool is close to saturation or the comprehensive evaluation index reaches the set threshold, the number of RPA robots running simultaneously is reduced, so that the third RPA robot and / or the second RPA robot stops working to reduce competitive occupation of resources and performance degradation.

[0059] (3) 60 data items are gradually added to the queue at a rate of 3 data items per second. The amount of data that triggers the first RPA robot job is 11. The amount of data that triggers a new RPA robot job is 10, which means that a new RPA robot job will be triggered every 10 data items added. The maximum number of RPA robot jobs that can run simultaneously is 3 by default.

[0060] At the 4th second, the number of data in the queue reaches 12. The first RPA robot starts the job and processes one data in the queue per second. Currently, there are 2 RPA robots waiting to run. The remaining data is 60-11=49, and 49 / 10=4.9>3.

[0061] According to the default maximum number of RPA robot jobs allowed to run simultaneously is 3, then at the 7th second, the data in the queue reaches 21, the second RPA robot starts the job, and the number of RPA robots currently waiting to run is 1; at the 11th second, the data in the queue reaches 33, and the third RPA robot starts the job. There is currently no RPA robot waiting to run; at the 20th second, the data is added, and there are 22 data left. The three RPA robots need another 8 seconds, and the total time is 28 seconds.

[0062] Since ((NX) / Z)>Y, the maximum number of RPA robot jobs allowed to run simultaneously is increased, and Y is changed to 5, satisfying ((NX) / Z)≤Y;

[0063] At this time, at the 7th second, the number of data in the queue reached 21, and the second RPA robot started the job; at the 11th second, the number of data in the queue reached 33, and the third RPA robot started the job; at the 14th second, the number of data in the queue reached 42, and the fourth RPA robot started the job; at the 17th second, the number of data in the queue reached 51, and the fifth RPA robot started the job; at the 20th second, the data was added, and there were 13 data left. The 5 RPA robots needed another 3 seconds to complete the processing, taking a total of 23 seconds.

[0064] The Y value is changed to 5. After the first RPA robot runs a job, if the resource utilization rate is lower than the set threshold, the remaining RPA robots are started in advance.

[0065] After the first RPA robot runs a job, if the RPA robot job response time exceeds the set threshold, the number of RPA robot jobs running simultaneously will be gradually increased until the new job triggering threshold is reached.

[0066] After 2-5 RPA robots are running jobs, when it is monitored that the resource pool is close to saturation or the comprehensive evaluation index reaches the set threshold, the number of RPA robots running simultaneously is reduced to reduce competitive resource occupation and performance degradation.

[0067] (4) 60 data items are gradually added to the queue at a rate of 3 data items per second. The amount of data that triggers the first RPA robot job is 31. The amount of data that triggers a new RPA robot job is 10, which means that a new RPA robot job is triggered every time 10 data items are added. The maximum number of RPA robot jobs that can run simultaneously is 3 by default.

[0068] That is, when the amount of data in the queue reaches 31, the first RPA robot job is started. At the 11th second, the amount of data in the queue reaches 33. The first RPA robot starts the job and processes one data in the queue per second. There are 60-31=29 data left, 29 / 10=2.9<3. The default maximum number of RPA robot jobs allowed to run simultaneously is not changed. The current number of RPA robots waiting to run is W=2.

[0069] If the threshold of W is set to 2, when the threshold of W is reached, the Z value is reduced to speed up the triggering of new RPA robot jobs, such as changing the Z value from 10 to 5;

[0070] At the 12th second, the number of data in the queue reaches 36, and the second RPA robot runs the job, which triggers the second RPA robot 2 seconds earlier than when Z=10.

[0071] At the 14th second, the queue has 42 pieces of data, and the third RPA robot runs the job;

[0072] At the 20th second, the data addition was completed. At this time, there were 37 data items left in the queue. It would take another 13 seconds to complete the data processing, with a total time of 33 seconds, which was 1 second earlier than when Z=10.

[0073] (5) 60 data are added to the queue at one time; the amount of data that triggers the first RPA robot job is 11; the amount of data that triggers a new RPA robot job is 10, that is, every 10 data added will trigger a new RPA robot job; the default maximum number of RPA robot jobs allowed to run simultaneously is 3;

[0074] By default, the maximum number of RPA robot jobs allowed to run simultaneously is 3. At this time, all three RPA robots start jobs. Each RPA robot processes 1 data per second, and it takes 20 seconds to complete the processing.

[0075] Since ((NX) / Z)=(60-11) / 10=4.9>3, Y becomes 5. The RPA operation system tries to run 5 RPA robots to satisfy ((NX) / Z)≤Y. At this time, it takes 12 seconds to complete the processing of the data in the queue.

[0076] This embodiment is only a further explanation of the present invention, not a limitation of the present invention. After reading this specification, those skilled in the art may make non-creative modifications to this embodiment as needed, but as long as it is within the scope of the claims of the present invention, it will be protected by the patent law.

Claims

1. A robotic process automation scheduling method based on queue processing efficiency, characterized in that: The method includes: Step 1: Collect the data to be processed and store it in a queue; Step 2: Check the data stored in the queue through the RPA task scheduling engine, and obtain the data to be processed in the queue based on the status and priority of the data in the queue; Step 3: Dynamically adjust the number of RPA robot jobs started, and the RPA robot calls the automated process to process data; Step 4: After the RPA robot completes the task, it generates data processing results and stores them in the queue.

2. A robotic process automation scheduling method based on queue processing efficiency according to claim 1, characterized in that: The amount of data in the queue is N, the data amount threshold for triggering the first RPA robot job is X, the data amount threshold for triggering a new RPA robot job is Z, and the maximum number of RPA robot jobs allowed to run simultaneously is Y by default; When ((NX) / Z)>Y, increase the maximum number of RPA robot jobs allowed to run simultaneously.

3. A robotic process automation scheduling method based on queue processing efficiency according to claim 2, characterized in that: The number of RPA robots currently waiting to run is W. A comprehensive evaluation index is created. The data volume threshold Z that triggers new RPA robot jobs is automatically set based on the current data volume waiting to be processed and the comprehensive evaluation index. When W reaches the set threshold, the Z value is reduced. When the comprehensive evaluation index is less than the set threshold, the Z value is increased. The comprehensive evaluation index is calculated according to a weighted combination of the RPA robot job execution efficiency, the resource utilization rate of the resource pool, and the RPA robot job response time, and the sum of the weights is 1.

4. The method for scheduling robotic process automation based on queue processing efficiency according to claim 3, characterized in that: The weight of the RPA robot job execution efficiency is higher than the resource utilization and the RPA robot job response time.

5. The method for scheduling robotic process automation based on queue processing efficiency according to claim 4, characterized in that: Comprehensive evaluation index = 0.4×RPA robot job execution efficiency + 0.3×resource utilization + 0.3×RPA robot job response time.

6. A robotic process automation scheduling method based on queue processing efficiency according to claim 2, characterized in that: When a new RPA robot job is triggered, the current status of the resource pool and comprehensive evaluation indicators are monitored to dynamically adjust the number of RPA robot jobs running simultaneously; When resource utilization is lower than the set threshold, increase the number of RPA robot jobs running simultaneously; When the RPA robot job response time exceeds the set threshold, increase the number of RPA robot jobs running simultaneously; When the resource pool is close to saturation or the comprehensive evaluation index reaches the set threshold, reduce the number of RPA robots running simultaneously.

7. The method for scheduling robotic process automation based on queue processing efficiency according to claim 1, characterized in that: When storing data to be processed in a queue, data is added to the queue gradually or in batches.

8. The method for scheduling robotic process automation based on queue processing efficiency according to claim 1, characterized in that: Each piece of data stored in the queue has an identification field, data priority, data status and data processing result.

9. The method for scheduling robotic process automation based on queue processing efficiency according to claim 8, characterized in that: The data status includes at least one of pending, being processed, processing exception, and business exception.

10. The method for scheduling robotic process automation based on queue processing efficiency according to claim 9, characterized in that: The status of the data in the queue changes according to the result of the RPA robot job execution.

Images