High Performance Fault Tolerant Missile Control System

Future missiles will be required to handle many diverse missions. These missions will require increased speed, precision, and survivability under a wider variety of flight conditions than current missiles. Therefore the role of modern flight control systems (FCS) must be expanded well beyond the traditional functions of augmenting stability and controllability with very limited control authority. IAI has already designed and implemented a functioning learning control system. The purpose of the work herein proposed is to evaluate the impact of our learning control system approach (called a Fuzzy CMAC (Cerebellar Model Arithrnetic Computer)) in the event of selected minor and major control system failures or actuator saturations. There are two issues. First, to what extent can our learning control system approach gracefully and automatically alleviate the impact of minor failures, and second, to what extent can the Fuzzy CMAC augment the integration of different non-collocated actuators and sensors and reconfigure the control systems in the event of major failures of certain actuators or saturated actuators. One major advantage of Fuzzy CMAC is that expert knowledge in terms of linguistic rules can be easily incorporated into the scheme. This knowledge is crucial to the successful operation of the missiles since different control effectors may have different control perforrnance. To activate or to inhibit certain control effectors under certain failure conditions will be of prime importance to missile flight control. The range of applications of the Fuzzy CMAC is very large. Other applications include medical monitoring, in-process control of manufacturing processes, real-time error compensation for machine tools, image classification, engine control to optimize performance for current environment (hot, cold, high altitude) and system health monitoring.