Adaptive Flight Control of an F-16 Without Prior Modeling

This paper presents an adaptive 6DOF flight-control system for an F-16. A nested-loop, multirate, feedback-control architecture is developed, and its performance is evaluated through numerical simulation. All controllers within the flight-control system use predictive cost adaptive control, which is an adaptive control technique that combines recursive-least-squares-based closed-loop, continual-online system identification with output-feedback model predictive control. The first contribution of the paper is the development of a command-control-conversion algorithm that converts the sequence of controls generated by an outer-loop model-predictive-control-based controller to a sequence of commands for an inner-loop model-predictive-control-based controller, where the controllers are updated at different rates. The second contribution is a complementary paradigm to traditional flight-control system design, which circumvents the need for prior modeling and control design. The predictive-cost-adaptive-control-based flight-control system operates in three modes, where each mode enables the ability to follow a different set of pilot commands, namely, rate, angle, and guidance commands. Numerical examples are presented to illustrate the performance of a predictive-cost-adaptive-control-based flight-control system, which uses no prior modeling information. The performance of this adaptive flight-control system is evaluated in the presence of ambient wind, sensor noise, unknown measurement gain and delay, and unknown payload changes.