Identification and quantification of concurrent control chart patterns using extreme-point symmetric mode decomposition and extreme learning machines

Abstract

Control chart pattern recognition (CCPR) is an important issue in statistical process control because unnatural control chart patterns (CCPs) exhibited on control charts can be associated with specific causes that adversely affect the manufacturing processes. In recent years, many machine learning techniques [e.g., artificial neural networks (ANNs) and support vector machines (SVMs)] have been successfully applied to CCPR. However, such existing research for CCPR has mostly been developed for identification of basic CCPs. Little attention has been given to the utilization of ANNs/SVMs for identification of concurrent CCPs (two or more basic CCPs occurring simultaneously) which are commonly encountered in practical manufacturing processes. In addition, these existing research for CCPR cannot provide more detailed CCP parameter information, such as shift magnitude, trend slope, cycle amplitude, etc., which is very useful for quality practitioners to search the assignable causes that give rise to the out-of-control situation. This study proposes a hybrid approach that integrates extreme-point symmetric mode decomposition (ESMD) with extreme learning machine (ELM) to identify typical concurrent CCPs and in addition to accurately quantify the major CCP parameter of the specific basic CCPs involved. The numerical results indicate that the proposed model can effectively identify not only concurrent CCPs but also basic CCPs. Meanwhile, the major CCP parameter of the identified concurrent CCP can also be accurately quantified.