Unwanted noise and vibration control using finite element analysis and artificial intelligence

https://doi.org/10.1016/j.apm.2013.10.039Get rights and content
Under an Elsevier user license
open archive

Abstract

A mechatronic approach integrating both passive and active controllers is presented in this study to deal with unwanted noise and vibration produced in an automobile wiper system operation. Wiper system is a flexible structure with high order, nonlinear model that is considered as a multi objective control problem since there is a conflict between its functionality quality in wiping and generated unwanted noise and vibration. A passive control technique using multi body system (MBS) model and finite element analysis (FEA) is introduced to identify the potential of the effectiveness of the physical parameters of wiper blade to give appropriate range to reduce the unwanted noise and vibration in the system. While, the significant contribution of active controller is to deal with uncertainties exerted to system within wiper operation. In passive control stage, natural frequencies of a uni-blade automobile wiper are determined using modal testing. Later, a 3-dimensional model of a wiper blade assembly is developed in multi body system design to investigate the good validation test and agreement of the physical behavior of the system in experiment with finite element analysis. Parametric modification via complex eigenvalue is adopted to predict instability of the wiper blade. In active control level, experimental data collected from the wiper system during its operation. A system identification model named nonlinear auto regressive exogenous Elman neural network (NARXENN) is developed for implying the active controller. A bi-level adaptive-fuzzy controller is brought in whose parameters are tuned simultaneously by a multi objective genetic algorithm (MOGA) to deal with the conflict interests in wiper control problem.

Keywords

Automotive wiper
Finite element analysis
Non-linear system identification
Fuzzy logic
Multi objective genetic algorithm

Cited by (0)