Combination Case-Based Reasoning and Clustering Method for Similarity Analysis of Production Manufacturing Process

Abstract

In production manufacturing process, the similarity analysis of production working status plays an important role in improving the economy and objectivity of management. It's very necessary to measure the similarity between each historical case and the target case of production working status to find the optimal working conditions. In this work, a similarity analysis methodology for production manufacturing process is proposed by using case-based reasoning and K-means clustering method. In order to improve K-means cluster efficiency, principal component analysis algorithm is taken to reduce feature attribute in original analysis space. In addition, the feature weighting of attributes is computed by deviation method in case-based reasoning system. Finally, the empirical research study is given to demonstrate that the evaluation results are more coincident with the reality and the proposed model's effectiveness.