Fault-tolerant control considering time-varying bounds on faults

Abstract

This paper presents a novel fault-tolerant control strategy to compensate the time-varying loss of actuators’ effectiveness. It considers intermediate situations where the fault is not determined precisely (unlike active approaches) but overall estimations about its rate and final value are available through the previous experiences and/or experiments. Based on the estimations, two upper and lower time-varying bounds on the actuators’ effectiveness are established to be exploited in the procedure of controller design. In a special case, where these bounds are constant, the method will be reduced to the conventional passive approach. Also, actuator saturation and the effects of [Formula: see text] disturbances are considered in the research. To tackle the conflict between attenuating the effect of L∞ disturbances and enlarging the domain of attraction, a linear matrix inequality optimization framework is suggested. The proposed method is implemented on a model of an unmanned aerial vehicle and the results are discussed.