Construction of the Time-Optimal Bounded Control for Linear Discrete-Time Systems Based on the Method of Superellipsoidal Approximation

Abstract

The speed-in-action problem for a linear discrete-time system with bounded control is considered. In the case of superellipsoidal constraints on the control, the optimal control process is constructed explicitly on the basis of the discrete maximum principle. The problem of calculating the initial conditions for an adjoint system is reduced to solving a system of algebraic equations. The algorithm for generating a guaranteeing solution based on the superellipsoidal approximation method is proposed for systems with general convex control constraints. The procedure of superellipsoidal approximation is reduced to solving a number of convex programming problems. Examples are given.