Phase compensation and anti-windup design for neural-aided sliding mode fault-tolerant autoland controller

Neural and Sliding Mode Control techniques are very popular for fault-tolerant control. However, control signals containing high frequency components will be present in these controllers due to rapid switching and high-gain adaptations involved. This leads to actuator rate and position saturations, and degrades the performance of the controllers. It is shown in this paper that by introducing suitable phase compensation and anti-windup schemes, the fault-tolerance performance of a neural-aided sliding mode auto-land controller can be improved.