Fuzzy learning systems for aircraft control law reconfiguration

Aircraft subsystem failures (e.g., actuator or sensor failures) or battle damage can cause catastrophic failures that can lead to loss of the aircraft. While experienced pilots can often compensate for failures, in certain emergency situations there is the need for computer-assisted or fully computer-automated reconfiguration of the aircraft control laws to save the aircraft. In this paper we show that the fuzzy model reference learning controller (FMRLC) can be used to reconfigure the nominal controller in an aircraft to compensate for various actuator failures without using explicit failure information (e.g., the time of the occurrence of the failure or its magnitude). After establishing a failure simulation testbed for the F-16 aircraft we introduce a new design procedure for the FMRLC that involves initializing the fuzzy controller so that it emulates the nominal control laws and viewing the “fuzzy inverse model” in the FMRLC as a fuzzy controller in the adaptation loop. Finally, we investigate the performance of the FMRLC for various failure conditions on an F-16 aircraft