On the Design of Optimal Input Signals in System Identification

The problem of designing optimal inputs in the identification of linear systems with unknown random parameters is considered using a Bayesian approach. The information matrix, which is positive definite for the class of systems analyzed, gives a measure of performance for the system inputs. The computation of the optimal closed-loop input mappings is shown to be a nontrivial exercise in adaptive control. Deterministic optimal inputs are shown to be easily computable. Numerical examples are given. A Kalman filter is used to estimate the parameters. A necessary condition for the Kalman filter not to diverge when applying linear feedback, is also given. (NASA-CE-137383) ON THE DESIGN OF I74-20171 1OPTIMAL INPUT SIGNALS IN SYSTEM IDENTIFICATION (Massachusetts Inst. of Tech.) 27 p HC $4.50 CSCL 12A Unclas G3/19 16921 * This research was conducte. in the Decision and Control Sciences Group of the M.I.T. Electronic Systems Laboratory with support e~tended by NASA Ames Research Center under grant NGL-22-009-124. ** Room 35-308, M.I.T., Cambridge, Massachusetts 02139. https://ntrs.nasa.gov/search.jsp?R=19740012058 2017-09-12T18:25:14+00:00Z