An extension of predictor control for systems with control time-delays

Publisher Website
Google Scholar
Abstract
The control of systems having time-delays in their control paths is considered and feedback laws are developed. A linear system is proposed and a quadratic criterion is optimized over an infinite interval to obtain time-invariant laws. A particularly simple law obtained by the solution of a non-delayed problem can be developed when all controls are delayed by the same amount, but when each control is delayed by a different amount the solution becomes quite complicated. Nevertheless this paper shows how solutions can be obtained by solving a sequence of tracking problems. In contrast to the state-delayed problem, where non-linear partial differential equations need be solved, only ordinary differential equations with single-point boundary values need be solved for the control-delayed problem. The time-invariant feedback law obtained involves a knowledge of the time-delay storages and performs a linear operation on these. The kernels of this operation cannot in general be derived as impulse responses of lumped systems.

This publication has 5 references indexed in Scilit: