Method for knowledge extraction and evolution in machine parts processing technological procedure

Abstract

The present invention provides a method for knowledge extraction and evolution in a machine parts processing technological procedure. The method comprises: building a machine parts information description model, arranging corresponding technological procedure data, and uniformly expressing the model and the data by adopting an XML technology; establishing a snowflake model process case base, and mapping the data in XML into a corresponding table; establishing a data mining system to perform clustering according process names and extract common process words in the technological procedure; complementing the first clustering of parts process cases according to parts information, performing clustering according to a technological route, extracting a typical process sequence and a typical structure process in the technological procedure, and recommending a typical process template; reusing process knowledge in the common process words, the process sequence, the typical structure process and the typical process template based on a case reasoning technology and by means of a part structure decomposition policy and a smart prompting technology; and establishing a memory and forgetting model for realizing evolution and essence extraction of the process knowledge. According to the method provided by the present invention, the reuse of the process knowledge in different granularity levels and the standardization of the process design are achieved.

Description

technical field [0001] The invention relates to a method for extracting part processing technology knowledge. In particular, it relates to a method for knowledge extraction and evolution in machine tool parts machining process regulations. Background technique [0002] In order to quickly respond to the needs of design and manufacturing, the design methods based on the modular configuration design and variant design of the product series are mostly used in the machine tool design, and the redesign is supplemented by the design means, so that the parts processing procedures accumulated in the past have greater reference and reliability. Reusability, but also faces the problem of rich data and poor knowledge. [0003] At present, the extraction of knowledge in process regulations mainly uses algorithms such as clustering, association rules, sequential pattern discovery, and rough sets to analyze and mine process information. Although the extracted knowledge has certain genera...

AI Technical Summary

Problems solved by technology

[0003] At present, the extraction of knowledge in process regulations mainly uses algorithms such as clustering, association rules, sequential pattern discovery, and rough sets to analyze and mine process information. Although the extracted knowledge has certain generality and typicality, it is fuzzy. It simplifies the consistency between the knowledge reuse angle and the extraction angle, which to a certain extent makes the extracted knowledge unreasonable and difficult to use.

At the same time, with the change of the production environment and the improvement of process technology, a large amount of useless process knowledge has been accumulated. Due to the lack of corresponding intelligent management, it has also increased the difficulty of knowledge reuse in the subsequent process.

Images