A new technology has been developed for improving performance and stability of control systems. This method represents a significant advancement in the state-of-the-art of adaptive control technology. The present invention is a new type of adaptive control law, called optimal control modification, which blends two control technologies together: optimal control and adaptive control.
The key features that differentiate this invention from the conventional art are: (1) the introduction of a damping mechanism that is proportional to a property known as persistent excitation to improve robustness of adaptation in the presence of persistently exciting signals, (2) the existence of linear asymptotic properties that make the method well suited for design and analysis for stability guarantee, and (3) the use of this adaptive control technology with a new modification of time-varying adaptive gain from two present methods — normalization and covariance adjustment — to further improve stability of the control systems in the presence of time delay and unmodeled dynamics. The method has gone through a series of validation processes ranging from many aircraft flight control simulations to a pilot-in-loop evaluation in the high-fidelity Advanced Concept Flight Simulator at NASA Ames that culminated in a successful flight test program on a NASA F/A-18A research test aircraft.
The invention provides a robust modification to the standard model-reference adaptive control based on an optimal control formulation. The modification enables fast adaptation for improving tracking performance without sacrificing stability robustness. The optimal control modification adaptive law can be tuned using a modification parameter to provide a trade-off between tracking performance and stability robustness. Increasing improves robustness to time delay and/or unmodeled dynamics. Simulation results for a wide variety of different aircraft models, as well as experimental data, demonstrate the effectiveness of the optimal control modification adaptive law, which improves the tracking performance significantly using a much larger adaptive gain than that for the standard model-reference adaptive control.