Control of Vehicle Suspension Using Neural Network

Bibliographic Information

Other Title
  • ニューラルネットワークを用いたアクティブサスペンションの制御
  • ニューラル ネットワーク オ モチイタ アクティブサスペンション ノ セイギョ

Search this article

Abstract

Holographic neural network, which is new algorithm of neural network, is applied to the control of a vehicle suspension. For the simplified suspension model, using the feedback gain obtained from the optimal control theory, the equation of motion is solved by Runge-Kutta method. The displacement, velocity and control force obtained is adopted as training data for neural network. For simulation the road displacement is assumed to be sinusoidal and its frequency is changed from 1.0Hz to 4.0Hz by 0.1Hz intervals. As the trained network is tested, the results for the road frequency not contained in the training data agree well with the results obtained from the optimal control theory.

Journal

Details 詳細情報について

Report a problem

Back to top