Robust LQR with actuator failure control strategies for 4DoF model of unmanned bicycle robot stabilised by inertial wheel

Abstract

The paper presents analysis of two control strategies, i.e. LQR and LQI, in application to control of continuous-time nonlinear system on the basis of its linearised discrete-time model, taking robustness approach into account. The applied control laws assure robustness subject to actuator failure what has been used to obtain better control quality (in the sense of selected performance indexes) in comparison to control laws that take no uncertainty into account. The uncertainty resembles here actuator failure or mismodeling case, whenever applied control signal results in different behaviour of the closed-loop system than desired, as in actuator failure case alone.