Determining the Optimum Process Mean for the Linear Transformation System

Abstract

The classical Taguchi quality model only considers the control of quality. Huang[9] proposed a model that took both quality and cost into account. Huang's model presents a trade-off problem between quality and cost, in which higher quality is not the final aim but the profit. In Huang's model, the symmetrical quadratic quality loss function was employed to stand for the loss of profit. In this paper, the process mean for the linear transformation is determined based on three types of loss functions: the symmetrical quadratic, asymmetrical quadratic, and asymmetrical linear loss functions. In addition, the application of Huang's model is extended to the process with constant coefficient of variation and the auto-correlated process. Some numerical examples are given for illustration. Keywords: Quality loss function; linear transformation system; mean; standard deviation.