Active Engine Noise Cancellation Using Vibro-Acoustic Modeling

Abstract

This thesis presents the implementation of narrowband (engine order noise target) Active Noise Control (ANC) systems in automotive environments via predictive analysis. This has become necessary to address the rising desires of auto manufacturers to shift further away from development on multiple prototype vehicles and instead towards a completely virtual first phase based on system modeling. This thesis demonstrates that the primary missing element barring a completely virtual ANC system model is the lack of an accurate secondary path estimation (often simply called an impulse response or IR). Included herein is a brief overview of the fundamental technologies and techniques used for much of the ANC industry. A thorough breakdown of the merits of various modeling assumptions with corresponding performance metrics and all coding used to generate the requisite secondary path response relationships. This study concludes that the described methods are fully capable for use with future prototype OEM models to evaluate the integration of narrowband ANC systems. A list of various future research that is made viable through this study is also included.