Crowdsourcing acquisition system and method based on motor fault knowledge

Abstract

The invention provides a crowdsourcing acquisition system and method based on motor fault knowledge, and belongs to the technical field of knowledge maps. The system comprises a data acquisition module which collects answers of experts through man-machine interaction and completes a crowdsourcing acquisition task. The man-machine interaction part is mainly divided into three modules: an applet part, a fault questionnaire part and a paper knowledge extraction part. The model training module uses a BERT model and is mainly used for performing a named entity recognition task, extracting key entities from an article, matching the entities and constructing a motor fault knowledge graph. And the knowledge construction module is used for carrying out matching according to predicates and editing distances after system knowledge training, and enabling fault description to correspond to fault elements, fault solutions and fault reasons. Through the technology, a motor fault knowledge crowdsourcing knowledge acquisition system is successfully designed and realized, motor fault knowledge can be extracted from various file formats, and a motor fault knowledge base and a knowledge graph are constructed.

Description

technical field [0001] The invention belongs to the technical field of knowledge graphs, and in particular relates to a crowdsourcing acquisition method and system based on motor fault knowledge. Background technique [0002] With the continuous advancement and development of science and technology, the degree of automation of motors used in various fields has gradually increased, and the internal structure of motors has gradually become more complex. There are more and more types of generators for different scenarios. Including hydroelectric generators, turbine generators, etc., and there will be many different types of generators. The maintenance and diagnosis of the above-mentioned various motors is a difficult problem at present. At the same time, as an important power equipment in various production fields, the motor has the advantages of low price, relatively simple overall structure, and relatively reliable, and undertakes more than 80% of the kinetic energy output i...

AI Technical Summary

Problems solved by technology

The technology using BERT is proposed on the basis of the seq2seq model. There is also an encoder (encoding) mechanism and a decoder (decoding) mechanism. The process of each layer of encoder is to use the encoding output of the previous layer as the encoding input of the next layer. , and finally get the encoding results, and then use these encoding results as the input of the decoder to decode step by step. However, it is easy to lose some information during the decoding and encoding process, resulting in the failure to obtain better precision and recall.

Images