Sensor scheduling with selection of input experimental conditions for identification of distributed systems

An approach to sensor scheduling problem in the context of proper choice of both the experimental settings and sensor locations that will be used to identify unknown parameters of a distributed system is presented. Specifically, given a finite number of possible sites at which sensors may reside and imposing additional limits on the number of experimental runs, a scheduling policy for discrete sensors is determined so as to maximize a criterion based on the Fisher information matrix associated with the estimated parameters. An efficient computational scheme based on the branch-and-bound method is provided for the solution of the resulting combinatorial problem. Finally, the proposed technique will be illustrated with simulations on a sensor scheduling problem regarding a distributed parameter system describing the performance of a magnetic brake.