Answers are grounded in this paper and cite their sources.
We present and analyze a learning control algorithm for a class of nonlinear systems with linear input action. Sufficient conditions for the convergence of the trajectories to the desired trajectory are given. These conditions suggest a way to design a learning operator given a model of the system and provide a measure of how close the model should be to the actual system to guarantee convergence.
No takes yet. Share an insight, caveat, or question.
John Hauser (1987) studied this question.