Answers are grounded in this paper and cite their sources.
We present a new adaptive control scheme where the control is chosen both to regulate and to elicit information. This is done by minimizing the one-step-ahead expected squared output, subject to a constraint on the trace of the one-step-ahead information matrix. We also incorporate a covariance-adaptive feature in the algorithm. Simulation results are given to compare the performance of this method to that of the perturbation controller of Wieslander and Wittenmark.
No takes yet. Share an insight, caveat, or question.
Alster et al. (1974) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: