On the Choice of a System and Sensor Configuration for Optimal Linear Estimation.
CALIFORNIA UNIV IRVINE SCHOOL OF ENGINEERING
Pagination or Media Count:
A system and sensor configuration control problem is formulated for a class of linear stochastic systems corrupted by Gaussian noise. The criterion function is a scalar measure of the terminal or steady-state error covariance matrix for the optimal linear estimator of the state of the system. A set of necessary conditions for optimality of a configuration control sequence is derived. The time-invariant case is examined in detail. A scalar example is solved analytically and a numerical algorithm is presented for the vector case. The time-varying problem with linear constraints is considered and two numerical methods of optimization are presented. Author Modified Abstract
- Statistics and Probability