Title: Characterisation of multiple interior noise metrics and translation of the voice of the customer

Authors: Sang-don Lee

Addresses: General Motors, 4156 Morningdale Drive, Troy, MI 48085, USA

Abstract: Several metrics represent noise characteristics such as engine idle, acceleration, and rough and smooth road noise because many factors affect vehicle noise. Engineers have to juggle several metrics to satisfy heterogeneous customer needs. All important design variables, which are often well correlated, should be included in translating the Voice Of the Customer (VOC) because the noise performance depends on several subsystems that are interconnected. Due to the highly correlated noise metrics, Multiple Linear Regression (MLR) is severely limited to incorporate special circumstances. Therefore, co-varying structures among interior noise metrics and the relationship between the latent structure and VOC are analysed by applying Principal Component Analysis (PCA) and Principal Component Regression (PCR).

Keywords: customer voice; noise metrics; principal component analysis; PCA; multicollinearity; multiple linear regression; MLR; characterisation; noise performance; voice of the customer; interior noise; vehicle noise; noise measurement.

DOI: 10.1504/IJVNV.2006.012784

International Journal of Vehicle Noise and Vibration, 2006 Vol.2 No.4, pp.341 - 356

Published online: 14 Mar 2007 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article