Volume 04,Issue 03

Performance Evaluation of Semi-Active Vehicle Suspension System Utilizing a New Adaptive Neuro-Fuzzy Controller Associated with Wavelet Transform

Authors

Abbas Soltani, Shahram Azadi, Seung-Bok Choi, Ahmad Bagheri

Abstract
In this paper, a magneto-rheological (MR) damper-based semi-active vehicle suspension is proposed using a new adaptive neuro-fuzzy inference system (ANFIS) controller associated with the wavelet transform (WT). A quarter-car suspension model is adopted for thecontroller design. When the resonance phenomena occurs on vertical wheel motion, the semi-active suspension system (SASS) based on ANFIS with constant gains cannot be very effective. In order to overcome this difficulty, the scaling factor of ANFIS is tuned by powerspectral density (PSD) and WT methods. Simulation results on the random road input with highly transient phenomena show that the WT is more proper compared to the PSD method for analyzing the signals including transient characteristics and improving SASS performance.

Keyword: Semi-active vehicle, MR damper, ANFIS controller, Wavelet transform, Power spectral density.

PDF [ 1394.51 Kb ] | Endnote File | XML