Process Capability Analysis For Correlated Profile And Multivariate Quality Characteristics

Process capability analysis is a quality management method that often used in quality improvement activities.It is possible to measure the degree of actual manufacturing process capability to meet the desired expectation through process capability analysis,and the measured results can provide the engineers with the direction of quality improvement.In the process capability analysis,process capability index(PCI)is usually selected as an indicator to analyze the process capability.In the manufacturing process of complex and precise products,the key quality of a product(or a process)is often represented by a profile model,even with the correlated profile and multivariate quality characteristics.Due to the complexity of profile data and the correlation between profile data and multivariate quality characteristics data,it is difficult to analyze the process capability of these quality characteristics,and the traditional univariate and multivariate methods of capability analysis can not be directly applied to the analysis of profile data.Thus,it becomes more and more urgent that how to effectively measure and analyze the process capability of such quality characteristics by constructing the PCI.For simple linear profile data,some researches about process capability analysis methods were done by domestic and overseas scholars.However,there is a lack of process capability analysis methods for correlated profile and multivariate quality characteristics.In this thesis,a method was proposed to analyze the process capability of correlated profile and multivariate quality characteristics.The method is based on a multi-dimensional vector that consists of a prediction profile vector and multivariate.First,the prediction profile vector was constructed with the profile parameters as elements,and then we combined the prediction profile vector and multivariate into a new multi-dimensional vector,and the problem of the correlated profile and multivariate was transformed into multiple problems.Further,a PCI based on area volume ratio was used to measure and analyze the process capability of correlated profile and multivariate.Finally,the performance of the proposed method in analyzing process capability was evaluated by a simulation with different scenarios of process variance,and a case was used to study and analyze in this paper.The results of a simulation study and a case study both proved the effectiveness and availability of the proposed method.