Two ways of stabilizing systems with input delays

Systems with input delays are a simple and intriguing example of distributed parameter systems. Studied by many researchers these systems are better analyzed by special methods than by the general theory of distributed parameter control systems. The present paper considers two basic approaches in feedback stabilization of such systems. The first one is the delay approximation. The error introduced by the approximation is considered as an uncertainty that may be treated via a robust synthesis. The second one is based on a Smith predictor version used together with some compensator designed for the delayless system. If the compensator is the result of a robust synthesis, then the implementation error for the predictor system could be also viewed as an uncertainty that the robust compensator solves in a satisfactory way. Some simulation results are given. Copyright (C) 1998 LFAC.