Importance Measure of Error Sources in Function Generation Mechanism based on Expected Taguchi Quality Loss

Abstract

A method to measure the relative contribution of the error sources to the output error of function generation mechanism is proposed in this paper, which is called error importance measure (EIM). The error sources and the resultant output error are inevitable in function generation mechanisms. In order to obtain optimum output accuracy, we should concentrate the limited resources on the most responsible error sources. Thus, a method to measure the importance of the error sources is desired. Firstly, the total quality loss of function generation mechanism across the whole motion process is derived based on quadratic Taguchi quality loss function. Then, by means of Taylor series expansion, we decompose the quality loss into a finite number of fractions, indicating the individual and interaction contributions of the error sources. Thirdly, the improved EIM indices are defined and the properties of the proposed method are discussed. At last, we offer an application case to demonstrate the effectiveness of the developed EIM method.