DECENTRALIZED COMPUTATIONAL TECHNIQUES FOR THE OPTIMIZATION OF DISTRIBUTED SYSTEMS.

Decentralized computational techniques for the optimization of a wide class of distributed parameter systems with constrained inputs are developed. A spatial-time discretization procedure is considered to convert the problem in hand to a time-varying lumped discrete system. The control function is divided between a number of local controllers and a higher controller. By this technique the computational effort of the two-point boundary value problem is decreased and the problem solution is always feasible during computer iterations.