A TOPSIS based robust optimization methodology for multivariable quality characteristics

Abstract

TOPSIS (Technique for order preference by similarity to ideal solution) method, as a multiple attribute decision making (MADM) methodology, is often used for solving problems having multivariable quality characteristics. In conventional TOPSIS method, responses are usually transformed into loss of quality, and then normalized by the largest quality loss. However, the conventional method lacks of robustness for it only takes the mean value into consideration and ignores the variance of response. Based on the conventional TOPSIS methodology, the proposed method makes an improvement for the purpose of robust design by taking the mean value and variance of response into consideration simultaneously. The improved TOPSIS method is illustrated by an example from literature, and the result indicates its reasonableness and effectiveness.