ABSTRACT
In order to improve the path following capability of vehicles under disturbances, an active steering system based on Generalized Predictive Controller (GPC) is proposed, of which a Controlled Auto-Regressive Integrated Moving-Average (CARIMA) model is used as its inner model. Considering the complexity and uncertainty of vehicle parameters, a Recursive Least Square Method is applied to estimate the parameters in the CARIMA model. The controller uses the deviation between the vehicle actual and desired lateral positons as input and steering wheel angle as output. Based on input steering wheel angle from driver, the final corrected steering wheel angle is obtained. The controller performances are examined in Simulink-CarSim environment in some typical conditions, e.g. under strong lateral wind condition in standard double-lane-change, etc. Thus, the feasibility and effectiveness of the designed controller are proved theoretically.
