Optimal sensor system design for state reconstruction

Abstract

A method is developed to evaluate the relationship between cost and performance in sensor system design. The conflicting requirements of minimizing cost and maximizing performance are met by a tradeoff analysis. Sensor cost is set proportional to systematic precision (defined as the inverse of the systematic error variance), and system performance is chosen to be a measure of the Cramer-Rao lower bound on the unknown parameter estimation accuracies. Expressions for performance sensitivities are developed as well as the necessary conditions for an extremum of the optimal precision problem.