Zhang, M.; Cheng, W., and Guo, P., 2015. Intelligent recognition of mixture control chart pattern based on quadratic feature extraction and SVM with AMPSO.

Take control charts as the principal statistical process control tools. The recognition accuracy of abnormal control chart patterns (CCPs) directly influences the quality control of production process. (Most of the existing studies focus on the basic recognition of abnormal CCPs, but the observed process data could be of mixture patterns in real world, which consists of two or three basic patterns. This paper introduces a hybrid intelligent model for recognizing the mixture control chart pattern that includes three main aspects: feature extraction, classifier and parameter optimization. In feature extraction, statistical and sharp features of observation data are used as data input and principal component analysis as the second feature extraction to get the effective data of the classifier. A multi-class support vector machine (SVM) is applied for recognizing the mixture patterns. Finally, adaptive mutation particle swam optimization is used to optimize the SVM classifier by searching the best values of the parameters of SVM kernel function. The simulation results show the proposed algorithm with little features has better recognition accuracy compared with other methods.