Study on typical process route mining method based on multilevel longest common subsequence information entropy and intelligent clustering model

Abstract

ABSTRACT A large number of manufacturing cases are accumulated by manufacturing enterprises in the process of operation and development, and therefore, one of the most effective ways to improve manufacturing efficiency and support innovation is to reuse these case resources reasonably. In reality, the first problem to be solved is to determine the case resources with high reuse value potential, so as to realize the high-value reuse of manufacturing case resources. With the purpose of scientific determination of the reuse objects and improvement of the reuse flexibility, a typical process route mining method based on multilevel longest common subsequence (LCS) information entropy and intelligent clustering model is proposed in this paper. First, a similarity calculation method of machining process route based on multilevel (LCS) information entropy is proposed, which can more comprehensively and accurately evaluate the similarity of machining process. On this basis, a process route clustering model based on spectral clustering idea and particle swarm optimization-Kmeans clustering algorithm is proposed, which realizes the clustering of process routes as per the similarity; in the end, the typical and representative process routes in each cluster are extracted, and the typical process routes are mined for reuse. In the end, it shows that the method proposed in this paper can effectively mine high-value process reuse objects and then can further support manufacturing case reuse through three verification cases.