Robust parameter design for signal-response systems using improved grey-TOPSIS model

To solve the robust parameter design problem for signal-response systems with multiple responses, a Taguchi-based method combining grey relational analysis and Technique for Order Preference by Similarity to an Ideal Solution (TOPSIS) was proposed. Principal component analysis was used to solve the correlation problem existing among multiple responses by taking the strategy that only eliminates the correlation but not reduce the dimensions. The improved grey-TOPSIS model was established using the weighted combination of Euclidean distance and grey relational degree to replace Euclidean distance, and the grey relative closeness degree was defined as the robustness index, which could reflect the relative position and change trend of the response's data sequences simultaneously. The robust parameter design results could be obtained based on the factor effect analysis of the grey relative closeness degree. The applicability and effectiveness of the proposed method was demonstrated with two engineering examples. The comparison results showed that the proposed method could achieve a better robust solution. © 2020, Editorial Department of CIMS. All right reserved.