Intelligent matching system based on business operation characteristics and key processes
By constructing an intelligent matching system based on business operation characteristics, the problems of insufficient flexibility and disordered execution in power grid safety supervision operation plans have been solved. It has achieved dynamic optimization and automated identification of key processes, improving execution efficiency and traceability.
Patent Information
- Application Number
- CN202511692467.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-18
- Publication Date
- 2026-02-10
AI Technical Summary
The existing power grid safety supervision operation plan lacks standardized rules, resulting in insufficient planning flexibility, poor dynamic adaptability, inability to quickly respond to changes in business needs, frequent resource conflicts, lack of identification of key processes and risk points, low execution efficiency, and overlapping effects between multiple work plans, leading to execution disorder and difficulty in tracing the source.
By employing a data acquisition module, a business feature construction module, a process map generation module, a work plan generation module, and a monitoring module, and combining semantic analysis, critical path method, genetic algorithm, and reinforcement learning algorithm, an intelligent matching system is constructed to identify key processes and generate multi-objective work plans, and to monitor and adjust them in real time to improve collaborative efficiency.
It enables dynamic optimization of power grid safety monitoring operation plans and automated identification of key processes, improving plan execution efficiency, reducing resource conflicts, and ensuring the accuracy of process execution and the convenience of traceability.
Smart Images
Figure CN121504064A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of intelligent management methods, specifically relating to an intelligent matching system based on business operation characteristics and key processes. Background Technology
[0002] Existing power grid safety monitoring plans rely on manual experience and lack standardized rules, resulting in insufficient flexibility and poor dynamic adaptability. When business needs change (such as sudden tasks or resource adjustments), manually prepared plans cannot respond quickly, cannot be dynamically optimized through data modeling, and cannot dynamically adapt to changes in business needs and the coupling relationship between processes. This leads to low plan execution efficiency, frequent resource conflicts, and a lack of automated identification of key processes and risk points, making it impossible to achieve data-driven dynamic risk assessment and precise control.
[0003] Furthermore, in the safety supervision of the power grid, there are multiple overlapping work plans, which in turn affect each other. This mutual influence leads to disorder in the execution and tracking of subsequent procedures, resulting in inadequate execution and difficulties in subsequent traceability.
[0004] In the current digital management solutions for work plans in the field of power grid safety supervision, when identifying key processes through work plans, existing technologies use natural language processing (NLP) technology to match keywords in the business or extract simple rules. However, in power grid safety supervision operations, there are a large number of complex steps. Summary of the Invention
[0005] To address the aforementioned problems in existing technologies, this invention provides an intelligent matching system based on business operation characteristics and key processes. This system resolves the issues of overlapping work plans, mutual influence among these plans, and subsequent disorder in the execution and tracking of processes, leading to inadequate process execution and difficulties in traceability.
[0006] The objective of this invention can be achieved through the following technical solution: an intelligent matching system based on business operation characteristics and key processes, comprising: The data acquisition module is used to collect data, including historical work order data and multi-source business plan data. The multi-source business plan data includes order attributes, equipment status, and process constraints. The historical work order data includes task steps, risk points, and control standards; The business feature construction module is used to extract business feature vectors based on multi-source data of business plans using semantic analysis. The business feature vectors include risk features, resource features, and time-series features of multiple single-line business plans. The process graph generation module is used to construct a directed process graph based on historical work order data using the critical path method, and to input the business feature vector into the directed process graph to identify key processes; The work plan generation module generates multi-objective work plans based on the key processes using time-constrained programming and genetic algorithms, distributes the work plans to workers, and obtains the execution progress of the work plans. The monitoring module is used to compare the execution process with the job plan using a reinforcement learning algorithm to determine the completion rate of the job plan.
[0007] Preferably, it also includes a conflict resolution module, which is used to obtain the updated job plan after the business plan is adjusted, and combine the execution process of the job plan before the adjustment with the mixed integer linear programming algorithm to re-plan and generate the adjusted job plan.
[0008] Preferably, the conflict resolution module uses a mixed-integer programming algorithm to regenerate the adjusted job plan, including: Data acquisition includes acquiring the adjusted work plan, collecting the execution progress of the current work plan in real time, generating the unfinished processes of the current work plan, loading the critical path, process dependencies, and business target weights of the current work plan, and preprocessing the acquired data. Do the current job schedule and the job schedules for incomplete processes overlap in terms of resource usage and conflict in terms of execution time? Overlapping and conflicting job plans will be sorted using a weighted approach. Generate the adjusted work plan.
[0009] Preferably, constructing the directed graph of the process includes: Enter historical work order data; Use process mining algorithms to extract dependencies between processes from historical work order data; The theoretical critical path of the process is constructed using CPM, and the actual process path is compared to correct the deviation of the directed graph. In the directed graph of the process, the nodes represent the steps of the process, and the edges represent the dependencies between the steps of the process.
[0010] Preferably, the data acquisition module further includes preprocessing the acquired data, the preprocessing including: The acquired data is divided into structured data and time-series data, and the structured data and time-series data are preprocessed separately, including: Structured data: using ETL process to clean the data, using SQL to parse historical work order data and extract fields; Time-series data: aggregating equipment status indicators through sliding window. Add timestamps, device IDs, and work order tags to the acquired multi-source data of the business plan.
[0011] Preferably, the execution steps of the business feature construction module include: The accident frequency and the failure rate of the equipment corresponding to the business line plan are extracted from the historical work orders. The risk value of the business line plan is calculated using the weight value method, and the risk value is set as the risk feature. Based on the GNN algorithm, the resource feature dependencies in multiple business single-line plans are calculated respectively, wherein the resource is used as a node and the dependency relationship is used as an edge to construct a resource availability vector; The resources include an operator information database, equipment information that needs to be maintained in the business line plan, and information on the tools required. The equipment information includes the type and status of the equipment. By obtaining the earliest start time and latest finish time of multiple business line plans, the time span of each business line plan is calculated. The time weight value of each business line plan is allocated and calculated using a weighting algorithm based on the time span and the latest finish time. The time sequence characteristics of each business line plan are generated based on the time weight values.
[0012] Preferably, the reinforcement learning algorithms used by the monitoring module include rule engine algorithms and graph neural network dynamic prediction.
[0013] Preferably, the data acquisition module further includes: using a sliding window mechanism to filter noise from the acquired multi-source business plan data, and filling in the missing data based on the KNN algorithm.
[0014] The beneficial effects of this invention are as follows: This invention employs a data acquisition module to collect data, including historical work order data and multi-source business plan data; a business feature construction module to extract business feature vectors based on the multi-source business plan data using semantic analysis; a process graph generation module to construct a directed process graph based on historical work order data using the critical path method, and input the business feature vectors into the directed process graph to identify key processes; and a work plan generation module to generate multi-objective work plans based on key processes using time-constrained programming and genetic algorithms. Through semantic analysis, the critical path method, and intelligent optimization algorithms, a fully intelligent system is constructed, from business feature extraction to dynamic work plan adjustment, thereby improving the efficiency of multi-process collaboration. Attached Figure Description
[0015] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings.
[0016] Figure 1 This is a system structure block diagram of the present invention. Detailed Implementation
[0017] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided.
[0018] Please see Figure 1 This embodiment provides an intelligent matching system based on business operation characteristics and key processes, including: The data acquisition module is used to collect data, including historical work order data and multi-source business plan data. The data acquisition module also includes preprocessing of the acquired data, which includes: The acquired data was divided into structured data and time-series data, and preprocessing was performed on the structured data and time-series data respectively, including: Structured data: Use ETL processes to clean the data and use SQL to parse historical work order data and extract fields; Time-series data: Aggregates device status indicators through a sliding window; Add timestamps, device IDs, and work order tags to the acquired multi-source business plan data. Use a sliding window mechanism to filter noise from the collected multi-source business plan data and fill in the missing data based on the KNN algorithm. The business plan's multi-source data includes order attributes, equipment status, and process constraints. Here, the order refers to the power grid maintenance order. For maintenance work orders generated for equipment in the power grid that needs maintenance, the equipment status is "power grid maintenance equipment" and "equipment that needs maintenance." The process constraints to be considered when performing maintenance based on the maintenance work order and the equipment it points to are the process flow diagram (XML format), process priority, and resource exclusivity rules (e.g., equipment A can only process one work order at a time). The specific constraints are further defined based on the environment in which the power grid maintenance takes place.
[0019] The business feature construction module is used to extract business feature vectors based on multi-source data of business plans using semantic analysis. The business feature vectors include risk features, resource features, and time series features of multiple single-line business plans. The execution steps of the business feature construction module include: The accident frequency and the failure rate of the equipment corresponding to the business line plan are extracted from the historical work orders. The risk value of the business line plan is calculated using the weight value method, and the risk value is set as the risk feature. Based on the GNN algorithm, the resource feature dependencies in multiple business single-line plans are calculated respectively, wherein the resource is used as a node and the dependency relationship is used as an edge to construct a resource availability vector; Resources include an operator information database, equipment information requiring maintenance in business line plans, and information on the tools needed. Equipment information includes equipment type and status. By obtaining the earliest start time and latest finish time of multiple business line plans, the time span of each business line plan is calculated. Combining the time span and the latest finish time, a weighting algorithm is used to assign and calculate the time weight values of multiple business line plans. Based on the time weight values, the time sequence characteristics of each business line plan are generated and sorted according to the time weight values.
[0020] The "risk points" and "control standards" texts are vectorized using a BERT pre-trained model. Risk feature vectors are generated through MeanPooling, and business feature vectors are extracted from historical work order data. The process graph generation module is used to construct a directed process graph based on historical work order data using the critical path method, and to identify key processes by inputting business feature vectors into the directed process graph. The work plan generation module uses time-constrained programming and genetic algorithms to generate multi-objective work plans based on key processes, distributes the work plans to workers, and obtains the execution progress of the work plans. The monitoring module is used to compare the execution progress with the task plan using reinforcement learning algorithms to determine the completion rate of the task plan. The reinforcement learning algorithms here include rule engine algorithms and graph neural networks to dynamically predict the completion status of the task plan. Among these methods, graph neural networks are used to dynamically predict the completion status of the work plan: The business feature vector is input into a pre-trained graph neural network model, where the embedding of process nodes includes attributes such as remaining time and resource requirements, and the edge weights reflect the strength of process dependencies. The model aggregates neighborhood information through a message passing mechanism, outputs a progress prediction for a certain period of time in the future, and then compares the progress prediction result with the execution process of the feedback work plan. The Q-learning algorithm is then used to calculate the optimal response action, where the triggering condition is: The delay probability of the process steps predicted by the GNN dynamic trajectory is greater than 30%, and the specific probability can be set according to the specific power grid maintenance process.
[0021] After initiating local rescheduling, the monitoring module sends the delayed process steps to the terminal, initiating a readjustment of the terminal processes.
[0022] The conflict resolution module is used to obtain the updated work plan after the business plan is adjusted. Combining the execution progress of the previous work plan, it uses a mixed-integer linear programming algorithm to regenerate the adjusted work plan. This includes adjusting subsequent work plans obtained from the linear integer programming, and making corresponding adjustments to the work plan based on the updates to the subsequent plans. Data acquisition includes acquiring the adjusted work plan, collecting the execution progress of the current work plan in real time, generating the unfinished processes of the current work plan, loading the critical path, process dependencies, and business target weights of the current work plan, and preprocessing the acquired data. Do the current job schedule and the job schedules for incomplete processes overlap in terms of resource usage and conflict in terms of execution time? Overlapping and conflicting work plans are sorted using a weighted approach to generate adjusted work plans.
[0023] By employing semantic analysis, critical path analysis, and intelligent optimization algorithms, a fully intelligent system is constructed, encompassing everything from business feature extraction to dynamic work plan adjustment, thereby improving the efficiency of multi-process collaboration. In this embodiment, constructing the directed graph of the process includes: Input historical work order data and parse out path dependency information from the historical work order data; Use process mining algorithms to extract dependencies between processes from historical work order data; The theoretical critical path of the process is constructed using CPM, and the actual process path is compared to correct the deviation of the directed graph. In this process-oriented graph, nodes represent steps within a process, and edges represent dependencies between these steps. The process-oriented graph is used to identify nodes and edges within the process graph, thereby clarifying the relationships between key processes.
[0024] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.
Claims
1. A smart matching system based on business operation characteristics and key processes, characterized in that: include: The data acquisition module is used to collect data, including historical work order data and multi-source business plan data. The multi-source business plan data includes order attributes, equipment status, and process constraints. The historical work order data includes task steps, risk points, and control standards; The business feature construction module is used to extract business feature vectors based on multi-source data of business plans using semantic analysis. The business feature vectors include risk features, resource features, and time-series features of multiple single-line business plans. The process graph generation module is used to construct a directed process graph based on historical work order data using the critical path method, and to input the business feature vector into the directed process graph to identify key processes; The work plan generation module generates multi-objective work plans based on the key processes using time-constrained programming and genetic algorithms, distributes the work plans to workers, and obtains the execution progress of the work plans. The monitoring module is used to compare the execution process with the job plan using a reinforcement learning algorithm to determine the completion rate of the job plan.
2. The intelligent matching system based on business operation characteristics and key processes according to claim 1, characterized in that: It also includes a conflict resolution module, which is used to obtain the updated job plan after the business plan is adjusted, and combine the execution process of the job plan before the adjustment with the mixed integer linear programming algorithm to re-plan and generate the adjusted job plan.
3. The intelligent matching system based on business operation characteristics and key processes according to claim 2, characterized in that: The conflict resolution module uses a mixed-integer programming algorithm to regenerate the adjusted job plan, including: Data acquisition includes acquiring the adjusted work plan, collecting the execution progress of the current work plan in real time, generating the unfinished processes of the current work plan, loading the critical path, process dependencies, and business target weights of the current work plan, and preprocessing the acquired data. Do the current job schedule and the job schedules for incomplete processes overlap in terms of resource usage and conflict in terms of execution time? Overlapping and conflicting job plans will be sorted using a weighted approach. Generate the adjusted work plan.
4. The intelligent matching system based on business operation characteristics and key processes according to claim 1, characterized in that: Constructing the directed graph of the process includes: Enter historical work order data; Use process mining algorithms to extract dependencies between processes from historical work order data; The theoretical critical path of the process is constructed using CPM, and the actual process path is compared to correct the deviation of the directed graph. In the directed graph of the process, the nodes represent the steps of the process, and the edges represent the dependencies between the steps of the process.
5. The intelligent matching system based on business operation characteristics and key processes according to claim 1, characterized in that: The data acquisition module further includes preprocessing the acquired data, the preprocessing including: The acquired data is divided into structured data and time-series data, and the structured data and time-series data are preprocessed separately, including: Structured data: using ETL process to clean the data, using SQL to parse historical work order data and extract fields; Time-series data: aggregating equipment status indicators through sliding window. Add timestamps, device IDs, and work order tags to the acquired multi-source data of the business plan.
6. The intelligent matching system based on business operation characteristics and key processes according to claim 1, characterized in that: The execution steps of the business feature construction module include: The accident frequency and the failure rate of the equipment corresponding to the business line plan are extracted from the historical work orders. The risk value of the business line plan is calculated using the weight value method, and the risk value is set as the risk feature. Based on the GNN algorithm, the resource feature dependencies in multiple business single-line plans are calculated respectively, wherein the resource is used as a node and the dependency relationship is used as an edge to construct a resource availability vector; The resources include an operator information database, equipment information that needs to be maintained in the business line plan, and information on the tools required. The equipment information includes the type and status of the equipment. By obtaining the earliest start time and latest finish time of multiple business line plans, the time span of each business line plan is calculated. The time weight value of each business line plan is allocated and calculated using a weighting algorithm based on the time span and the latest finish time. The time sequence characteristics of each business line plan are generated based on the time weight values.
7. The intelligent matching system based on business operation characteristics and key processes according to claim 3, characterized in that: The reinforcement learning algorithms used in the monitoring module include rule engine algorithms and graph neural network dynamic prediction.
8. The intelligent matching system based on business operation characteristics and key processes according to claim 1, characterized in that: The data acquisition module also includes: using a sliding window mechanism to filter noise from the acquired multi-source business plan data, and filling in the missing data based on the KNN algorithm.