Multi-Objective Optimization of Vehicle Sound Package in Middle Frequency Using Gray Relational Analysis Coupled with Principal Component Analysis
shuming chen,dengfeng wang,tianze shi,jing chen
DOI: https://doi.org/10.1142/S0219477515500108
2015-01-01
Fluctuation and Noise Letters
Abstract:This research studies optimization design of the thickness of sound packages for a passenger car. The major characteristics indexes for performance determined to evaluate the process are sound pressure level of the interior middle frequency noise and weight of the sound package. Three kinds of materials of sound packages are selected for the optimization process. The corresponding parameters of the sound packages are the thickness of the insulation plate for outer side of the firewall, thickness of the sound absorbing wool for inner side of the firewall, thickness of PU foam for the front floor, and thickness of PU foam for the rear floor, respectively. In this paper, the optimization procedure is a multi-objective optimization. Therefore, gray relational analysis (GRA) is applied to decide the optimal combination of sound package parameters. Furthermore, the principal component analysis (PCA) is used to calculate the weighting values which are corresponding to multiple performance characteristics. Then, the results of the confirmation tests uncover that GRA coupled with principal analysis methods can effectively be applied to find the optimal combination of the thickness of the sound packages at different positions for a passenger car. Thus, the proposed method can be a useful tool to improve the automotive interior middle frequency noise and lower the weight of the sound packages. Additionally, it will also be useful for automotive manufactures and designers in other fields.