An inverse dynamics-based dynamic programming method for optimal control problems of linear systems

In solving optimal control problems, the conventional dynamic programming method often requires interpolations to determine the optimal control law. As a consequence, interpolation errors often degenerate the accuracy of the conventional dynamic programming method. In view of this problem, this paper introduces an inverse dynamic programming method to eliminate the interpolation requirement for systems with linear dynamics. Simulation results show that the proposed approach provides more reliable solutions than the conventional dynamic programming method.