A generalized inverse cascade method to identify and optimize vehicle interior noise sources

The noise, vibration and harshness (NVH) emitted by a vehicle are very important to a customer's perception of the vehicle quality. A vehicle's NVH can be improved by considering the three following facets: the noise source, transfer path, and receiver. The identification and optimization of vehicle interior noise sources is crucial when attempting to reduce noise levels and improve sound quality. Although traditional methods, such as those utilizing sound pressure levels, nearfield acoustic holography, and transfer path analysis, can provide the magnitudes and contributions of noise sources, they cannot present specific methods for optimizing those noise sources. This study proposes a new method, the generalized inverse cascade method (GICM), to solve this problem. The GICM combines systems engineering with the interval optimization technique to identify and optimize vehicle noise sources. Applying the GICM to a decision problem involves the following three steps: (1) constructing the decision problem as a cascade tree; (2) developing a numerical model to quantify the cascade tree; and (3) solving the numerical model using the interval optimization method. A Volkswagen sedan is used in this study as an example, and a vehicular road test and subjective evaluation are implemented to record and evaluate the interior noise. The GICM, identifies potential abnormal interior noise sources, and a modified method is presented to optimize the abnormal noise sources by calculating the feasible intervals of design variables. A verification experiment shows that the vehicle interior noise is successfully optimized, thereby validating the proposed GICM. (C) 2019 Elsevier Ltd. All rights reserved.