Citation

Barethiye V, Pohit G, Mitra A. Eng. Sci. Technol. Int J. 2017; 20(6): 1610-1622.

Copyright

(Copyright © 2017, Karabuk University, Publisher Elsevier Publishing)

DOI

10.1016/j.jestch.2017.12.003

PMID

unavailable

Abstract

Modeling dynamic characteristics of an automotive shock absorber is a challenging task due to its complex behavior. In the present paper, the nonparametric and hybrid approach is proposed to represent the nonlinear and hysteresis characteristics of the shock absorber. An experiment is carried out on a car damper utilizing INSTRON to obtain force-velocity characteristics of the shock absorber. The experimental data is used to devise two different models, namely, piecewise linear model and hysteresis model, to capture the damping properties of the absorber and for consequent use in simulations. The complexity involved due to certain physical phenomenon such as oil compressibility, gas entrapment etc. gives rise to hysteresis behavior and the present paper tries to model such behavior with the help of Neural Networks. Finally, a combined (hybrid) shock absorber model (including the characteristics of both piecewise linear and hysteresis behavior) is developed in Simulink and integrated into a quarter car simulation to verify its feasibility. The results generated by the combined (hybrid) model are compared with linear as well as piecewise linear model and the comparison shows that the proposed model substantially a better option to study the vehicle characteristics more accurately and precisely.

Language: en

Keywords

INSTRON 8801; Neural network; Nonlinear and hysteresis; Quarter car; Shock absorber modeling; Simulink