An equipment fault diagnosis method and system based on natural language processing and case-based reasoning

Abstract

The invention discloses an equipment fault diagnosis method based on natural language processing and case-based reasoning. The equipment fault diagnosis method comprises the following steps: establishing a fault diagnosis knowledge base; Establishing a fault diagnosis case library; extracting diagnosis keywords from the fault diagnosis case library, establishing a fault diagnosis word library, extracting feature tags, and calculating weight matrixes of the feature tags; obtaining the similarity between the target fault case and the source fault case through the weight matrix, and then obtaining the matching degree of the target fault case and the source fault case; Obtaining a source fault case similar to the target fault case through the matching degree, and regarding the source fault case with the matching degree greater than a preset threshold as a similar case; And then making a response and evaluation to an actually occurring fault, and maintaining the case library according to anevaluation result. Through the method and the system, maintenance personnel can be helped to quickly and accurately determine the part and the reason of the equipment fault after the equipment faultoccurs, so that the improvement of the yield and the economic benefit of the equipment in the same time period is embodied.

Description

technical field [0001] The invention relates to the technical field of artificial intelligence fault diagnosis, in particular to an equipment fault diagnosis method and system based on natural language processing and case reasoning. Background technique [0002] Modern automation equipment has production characteristics such as high speed, refinement, and complex processes. The equipment itself also has equipment characteristics such as a large number of parts, complex structures, and extremely high requirements for parts to cooperate with each other. [0003] Taking the packaging machine in the cigarette industry as an example, as a typical multi-stage transmission equipment, the failure probability of the packaging machine in actual production is very high and there are many causes, and its failures are diverse, random and coupled. Any part of the unit at any time may cause a failure. [0004] At this stage, for the diagnosis and maintenance of equipment during fault diag...

AI Technical Summary

Problems solved by technology

[0004] At this stage, for the diagnosis and maintenance of equipment during fault diagnosis, the empirical diagnosis of faults mainly relies on the past experience of on-site maintenance personnel, which takes a long time, has certain subjectivity, and is highly dependent on personnel technology. There is no clear and quantitative method to quickly identify and locate various faults of packaging machines

Therefore, those who are able to perform effective equipment maintenance are mostly experienced technicians and senior masters. It takes several years for new employees to be trained, proficient and accumulate technical experience. The inheritance of technical experience is poor and the inheritance efficiency is low.

[0005] Because it is often difficult to quickly and effectively locate the cause of the failure from the outside to the inside with solidified technical experience when the equipment fails and shuts down, it is also difficult to provide objective and efficient maintenance guidance in a very short time

As a result, during the period from the occurrence of the entire failure to the successful maintenance of the equipment, the equipment cannot work normally, which will cause a decrease in its effective operation rate, which will lead to a large loss of production efficiency, and may also cause a certain degree of primary and secondary equipment loss.

[0006] In addition, there are already some equipment fault diagnosis methods based on case reasoning, such as the invention patent "A Method for Fault Diagnosis of Steam Turbine Heater Based on Case Reasoning" (CN201711375932.4), which uses a formal concept to structure the knowledge required for case reasoning However, the structured description will inevitably lead to the incomplete match between the fault description and the actual situation, and there are certain limitations

Invention patent "a kitchenware fault diagnosis method based on case-based reasoning" (CN201110043539.1) avoids the incomplete and inconsistent structural description to a certain extent by semi-structured description of kitchenware faults, but numerically Indicates that there is no clear threshold limit for the severity of qualitative fault symptoms, and there is strong subjectivity

Images