Analysis of multivariate measurement system based on KPCA method
Measurement system analysis is an important content of statistical quality improvement. In multivariate measurement system analysis, most researchers ignore the impact of variation caused by environmental factors on the whole measurement system. However, in the actual production process, environmental factors are not constant, and some changes always occur. In this paper, Kernel Principal Component Analysis(KPCA) based multivariate measurement system analysis, combined with dimensional analysis of nonlinear environmental factor data to raise dimension first and then reduce dimension, multivariate variables reduced to one or two variables, using variance analysis to evaluate the capacity of the measurement system. Finally, this paper uses the data of a food manufacturing company in the production of milk tea companion solid beverage to evaluate the capability of the multivariate measurement system and verify the effectiveness of the method.
Keywords: Measurement system analysis; Environmental factors; Kernel Principal Component Analysis (KPCA); Dimensional analysis
