Abstract:
A regular method of designing an optimal law of control through prediction of the plant motion is analyzed and shown to lead to an iterative procedure of computing the control signal for each time. The extreme control is shown to be optimal in terms of a quadratic functional.