Retrospective Prediction Control with Respect to the Generalized-Action Criterion

摘要

In [1], A.A.Krasovskii pioneered the analysis of a control algorithm based on a generalized-action functional with the use of a retrospective model applied to the solution of the problem of external correlative navigation. In this paper, the well-known algorithm with a prediction model and a sensitivity matrix [2] is developed with regard to the above-mentioned result. This algorithm allows for the joint (performed during the operation of a dynamic object) synthesis of controls that are optimal in the sense of a generalized-action function. New versions of the algorithm are constructed on the basis of a new form of a generalized-action functional. This form explicitly takes into account the prehistory (retrospection) of an object's motion. An example is presented.