Estimate of nonlinear system of engine noise transfer based on higher order statistics (HOS)

Abstract

Based on the higher order statistics (HOS), a step variable adaptive method for the nonlinear system estimation of engine noise transfer is proposed. It gives consistent estimation of parameters of linear and nonlinear components separately, so every step is easy to explore. HOS makes this algorithm insensitive to any i.i.d. noise. The variable step of the adaptive algorithm gives fast convergence rate. Our experiments on a microbus show that the proposed approach gives a system function that precisely describes the transfer of engine noise.