Artificial neural network predication and validation of optimum suspension parameters of a passive suspension system

Nagarkar, Mahesh P., El-Gohary, M. A., Bhalerao, Yogesh J., Vikhe Patil, Gahininath J. and Zaware Patil, Rahul N. (2019) Artificial neural network predication and validation of optimum suspension parameters of a passive suspension system. Applied Sciences, 1 (6). ISSN 2523-3963

[img]
Preview
PDF (Accepted_Manuscript) - Submitted Version
Download (1MB) | Preview

Abstract

This paper presents the modeling and optimization of quarter car suspension system using Macpherson strut. A mathematical model of quarter car is developed, simulated and optimized in Matlab/Simulink® environment. The results are validated using test rig. The suspension system parameters are optimized using a genetic algorithm for objective functions viz. vibration dose value (VDV), frequency weighted root mean square acceleration (hereafter called as RMS acceleration), maximum transient vibration value, root mean square suspension space and root mean square tyre deflection. ISO 2631-1 standard is adopted to assess ride and health criterion. Results shows that optimum parameters provide ride comfort and health criterions over classical design. The optimization results are experimentally validated using quarter car test setup. The genetic algorithm optimization results are further extended to the artificial neural network simulation and prediction model. Artificial neural network model is carried out in Matlab/Simulink® environment and Neuro Dimensions. Simulation, experimental and predicted results are in close correlation. The optimized system reduces the values of VDV by 45%. Also, RMS acceleration is reduced by 47%. Thus, the optimized system improved ride comfort by reducing RMS acceleration and improved health criterion by reducing the VDV. Finally ANN can be used for predicting the optimum suspension parameters values with good agreement.

Item Type: Article
Faculty \ School:
Related URLs:
Depositing User: LivePure Connector
Date Deposited: 07 Feb 2020 05:09
Last Modified: 20 Nov 2020 00:50
URI: https://ueaeprints.uea.ac.uk/id/eprint/73988
DOI: 10.1007/s42452-019-0550-0

Actions (login required)

View Item View Item