Adaptive Identification for Heavy-Truck Stability Control

Abstract

SUMMARY In this paper a new procedure is proposed to predict the vehicle behaviour when the vehicle is approaching to some critical point (roll-over, jackknifing). The on-board computer automatically identifies some of the vehicle's parameters, and runs the vehicle model getting the information from the sensors installed on the vehicle. When the vehicle approaching to some stability limit the safety system is activated. An adaptive linear quadratic Gaussian (LQG) controller is calculated at each time step of the identification phase as an estimate of the plant parameters from which the parameters of the LQG controller are updated. Meanwhile the parameters of the controller are modified in order to get an adaptive robust controller with the attractive properties of the linear quadratic (LQ) controller in such a way that loop transfer recovery (LTR) of the original state feedback is achieved.