CN-121981700-A - Maintenance task scheduling management method, device, computer equipment and storage medium

CN121981700ACN 121981700 ACN121981700 ACN 121981700ACN-121981700-A

Abstract

The application relates to a maintenance task scheduling management method, a maintenance task scheduling management device, computer equipment and a storage medium. According to the method, fault description and task context information are automatically analyzed after a maintenance task is detected, multidimensional similarity matching is achieved through a structured case knowledge base, historical cases closest to the current task can be rapidly located, a recommended disposal scheme with experience guidance is generated, global optimization is further conducted on the scheme through combination of an optimized scheduling model, human resource loads are balanced, total completion time is shortened, a system dynamically monitors key resource states through real-time resource states and automatically resolves conflicts, and therefore an executable alternative scheme meeting resource constraint is output, finally a digital work card is generated through semantic matching and integrated into a maintenance plan, full-flow closed-loop optimization from task analysis and scheme generation to plan execution is achieved, and the intelligent level and execution accuracy of maintenance plan making are greatly improved.

Inventors

MA YUEZHOU
WANG YIJIE
ZHU QI
WANG RENJIE
XIE YANQI
ZHOU TAORAN
LI LI
WANG YISHENG
LIU DUN
ZHANG LEI

Assignees

上海船舶工艺研究所(中国船舶集团有限公司第十一研究所)

Dates

Publication Date: 20260505
Application Date: 20251219

Claims (10)

1. The maintenance task scheduling management method is characterized by comprising the following steps of: when an input maintenance task is detected, analyzing fault description and task context information of the maintenance task, carrying out multidimensional similarity matching through a structured case knowledge base, screening out a historical case with highest matching degree, and generating a recommended disposal scheme; Constructing an optimized scheduling model aiming at minimizing total finishing time and balancing team load, optimizing the recommended treatment scheme by using the optimized scheduling model according to a professional capability matrix of personnel, task priority and the upper limit number of parallel tasks as constraint conditions, and outputting a global optimized distribution scheme; Tracking idle, scheduled and occupied states and time lines of key resources by maintaining a real-time resource state machine, simulating and detecting resource conflict by the real-time resource state machine, and eliminating the resource conflict in the global optimization distribution scheme to generate an alternative scheme meeting the current resource state; Analyzing sub-task document fragments according to the alternative scheme, carrying out semantic similarity matching on the sub-task document fragments and nodes in a maintenance work card knowledge graph, and creating a digital work card for each sub-task document fragment according to a matching result; the currently executing digital work card is associated with the newly created digital work card to a maintenance plan and the maintenance plan is executed.
2. The maintenance task scheduling management method according to claim 1, characterized in that the maintenance task scheduling management method further comprises: uploading the grade repair task function and the original repair engineering list of the historical cases to a system in batches, and identifying the document types and storing the document types in a classified manner by the system to form a project basic document library; establishing character models in a system, and defining identity information, static operation authority and capability attribute sets containing professional skill labels and task load parameters of each character model; Based on the document content in the project basic document library, carrying out structural analysis on the document by utilizing a natural language processing mode, combining an analysis result with the character model, and establishing a maintenance knowledge graph to form a structural case knowledge base.
3. The maintenance task scheduling management method according to claim 2, further comprising, after establishing the maintenance knowledge graph: Invoking a maintenance knowledge graph and real-time resource state data, and performing feasibility assessment on a repair range in the original maintenance engineering bill; associating the character model of corresponding responsibility with the original repair order when a repair scope in the original repair order is viable; And when the repair range in the original repair engineering bill is not feasible, sending out infeasible evaluation prompt information.
4. The maintenance task scheduling management method according to claim 1, characterized in that the maintenance task scheduling management method further comprises: carrying out structured data acquisition on the result of executing the maintenance plan, and judging whether the task of the maintenance plan is completed or not; if yes, carrying out data-driven knowledge evolution on the structured case knowledge base; if not, continuing to execute the maintenance plan for the unfinished task.
5. The maintenance task scheduling management method according to claim 1, wherein the simulation of detecting resource conflicts by the real-time resource state machine and eliminating resource conflicts in the global optimal distribution scheme generates alternatives satisfying the current resource state, comprising: Splitting to form work cards according to the global optimization distribution scheme, abstracting all the work cards into a task dependency model, wherein the process sequence relation is represented in a directed acyclic graph form, nodes represent the work cards, and edges represent the process sequence and resource dependency relation; Constructing a resource dependency graph according to the resource occupation relationship, and detecting cyclic dependency and mutually exclusive resource competition in the resource dependency graph through static analysis of the directed acyclic graph; Based on the dynamic simulation of a time axis, detecting whether space-time overlapping conflict exists when the work card is estimated to work, and generating a resource conflict resolution suggestion by applying a graph rearrangement algorithm; and eliminating the resource conflict in the global optimization distribution scheme according to the resource conflict resolution suggestion to generate an alternative scheme meeting the current resource state.
6. The maintenance task scheduling management method according to claim 1, wherein the parsing of the sub-task document fragment according to the alternative scheme includes: carrying out semantic compliance examination on the subtask document fragments generated according to the alternative scheme, and carrying out semantic verification and format normalization detection on the contents of the subtask document fragments through a formatting rule engine; and when detecting that the content does not accord with the maintenance standard or the safety standard, generating a correction rule prompt and feeding back to the task editing interface.
7. The maintenance task scheduling management method according to claim 1, wherein the associating the currently executed digital work card with the newly created digital work card to a maintenance plan and executing the maintenance plan includes: When the digital work card is associated with the maintenance plan, dynamically generating a quality check point list and a security risk prompt card according to the task type, the risk level and the quality control requirement, and binding the quality check point and the security risk prompt in a task packaging stage; In the process of executing the maintenance plan, generating a field operation interface through dynamic driving of a currently executed digital work card, and restricting the format and semantic standard of input data through interactive modes of formalization, option and code scanning identification; And after the maintenance plan is executed, generating a key performance index chart of the task execution process according to the collected full-flow structured data, and forming an engineering summary report according to a standardized template.
8. A maintenance task scheduling management apparatus, the apparatus comprising: The task analysis module is used for analyzing fault description and task context information of the maintenance task when the input maintenance task is detected, carrying out multidimensional similarity matching through the structured case knowledge base, screening out a history case with highest matching degree and generating a recommended disposal scheme; The optimization scheduling module is used for constructing an optimization scheduling model aiming at minimizing total finishing time and balancing team load, optimizing the recommended treatment scheme by using the optimization scheduling model according to a professional capability matrix of personnel, task priority and the upper limit number of parallel tasks as constraint conditions, and outputting a global optimization distribution scheme; The resource coordination module is used for tracking the idle, preset and occupied states and time lines of key resources by maintaining a real-time resource state machine, simulating and detecting resource conflict by the real-time resource state machine, and eliminating the resource conflict in the global optimization distribution scheme to generate an alternative scheme meeting the current resource state; The work card generation module is used for analyzing sub-task document fragments according to the alternative scheme, carrying out semantic similarity matching on the sub-task document fragments and nodes in a maintenance work card knowledge graph, and creating a digital work card for each sub-task document fragment according to a matching result; And the plan execution module is used for associating the currently executed digital work card and the newly created digital work card to a maintenance plan and executing the maintenance plan.
9. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the steps of the method according to any one of claims 1 to 7 when the computer program is executed by the processor.
10. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the steps of the method of any of claims 1 to 7.

Description

Maintenance task scheduling management method, device, computer equipment and storage medium Technical Field The present application relates to the field of server technologies, and in particular, to a maintenance task scheduling management method, a device, a computer device, and a storage medium. Background The current solution adopted by ship grade repair is a method of combining static planning with manual management and control, namely, firstly, annual repair planning is formulated according to the service life and repair grade standard of the ship, single ship repair nodes and resource requirements are defined, then, units such as troops, repair factories, equipment suppliers and the like are coordinated through manual butt joint, repair progress is tracked by means of paper records or basic forms, problems are checked and plans are adjusted through regular conferences in the repair process, and finally, equipment delivery and post-repair summarization are completed. The existing digital work card-based system has preliminarily realized the digital management of maintenance tasks, but when the complex scene of 'time is tight, tasks are heavy and professionals cross many' of equipment level repair is dealt with, the static, stiff and passive cooperative mode becomes a new efficiency bottleneck. The prior art is mainly used for carrying out the grade maintenance of the ship related equipment based on a static plan combined with a manual management and control method, and has the main defects that the static plan is mainly used, the influence of sudden tasks on the repair plan is not considered, the maintenance resource allocation depends on manual statistics, digital support is lacking, risk identification is not comprehensive, and countermeasures are not effective. Disclosure of Invention Based on the method, the device, the computer equipment and the storage medium for scheduling and managing the maintenance tasks are provided, and the method, the device and the computer equipment are used for solving the technical problems that the existing digital work card mode is a static plan combined with a manual management and control method for performing the level maintenance of the ship related equipment, so that the task burst is difficult to happen, and the data loss and the risk are high. In one aspect, a maintenance task scheduling management method is provided, the method includes: when an input maintenance task is detected, analyzing fault description and task context information of the maintenance task, carrying out multidimensional similarity matching through a structured case knowledge base, screening out a historical case with highest matching degree, and generating a recommended disposal scheme; Constructing an optimized scheduling model aiming at minimizing total finishing time and balancing team load, optimizing the recommended treatment scheme by using the optimized scheduling model according to a professional capability matrix of personnel, task priority and the upper limit number of parallel tasks as constraint conditions, and outputting a global optimized distribution scheme; Tracking idle, scheduled and occupied states and time lines of key resources by maintaining a real-time resource state machine, simulating and detecting resource conflict by the real-time resource state machine, and eliminating the resource conflict in the global optimization distribution scheme to generate an alternative scheme meeting the current resource state; Analyzing sub-task document fragments according to the alternative scheme, carrying out semantic similarity matching on the sub-task document fragments and nodes in a maintenance work card knowledge graph, and creating a digital work card for each sub-task document fragment according to a matching result; the currently executing digital work card is associated with the newly created digital work card to a maintenance plan and the maintenance plan is executed. In one embodiment, the maintenance task scheduling management method further includes: uploading the grade repair task function and the original repair engineering list of the historical cases to a system in batches, and identifying the document types and storing the document types in a classified manner by the system to form a project basic document library; establishing character models in a system, and defining identity information, static operation authority and capability attribute sets containing professional skill labels and task load parameters of each character model; Based on the document content in the project basic document library, carrying out structural analysis on the document by utilizing a Natural Language Processing (NLP) mode, combining an analysis result with the character model, and establishing a maintenance knowledge graph to form a structural case knowledge base. In one embodiment, after establishing the maintenance knowledge graph, the method further comprises: Invoking a mainte