Neural networks representation of a vehicle model: 'Neuro-Vehicle (NV)' Online publication date: Wed, 28-May-2014
by A. Ghazizadeh, A. Fahim, M. El-Gindy
International Journal of Vehicle Design (IJVD), Vol. 17, No. 1, 1996
Abstract: Recent developments in the area of the artificial neural networks (NN) provide an alternative approach to the modelling of vehicular dynamics, particularly near their operational limits where the system is highly nonlinear. The objective of this paper is to investigate the ability of a NN called the 'Neuro-Vehicle' to simulate the dynamic behaviour of a two-axle vehicle. The input to the vehicle system is the forward speed and the steering angle, and the output responses are the yaw rate, the lateral acceleration, and the lateral load transfer ratios. To enhance the NN performance, current and past status of the Neuro-Vehicle are fed back to the network. A back-propagation learning rule with adaptive learning and momentum was used to teach the NN. The training data were produced using a simplified non-linear vehicle model with a quasistatic load transfer assumption. These data pertain to three manoeuvres each carried out at five different speeds. The performance of the Neuro-Vehicle shows that neural networks are fast and relatively accurate in predicting the nonlinear behaviour of a vehicle. Since the approach is fairly new and in its early stage, some problems associated with this study will be discussed and attempts to solve them will appear in future work. The work described in this paper is the first stage of a long-term project aimed at developing an Intelligent Rollover and Stability Enhancement Warning System.
Online publication date: Wed, 28-May-2014
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.Complimentary Subscribers, Editors or Members of the Editorial Board of the International Journal of Vehicle Design (IJVD):
Login with your Inderscience username and password:
Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.
If you still need assistance, please email email@example.com