Nonlinear stabilization through long input delay—Part II: The strict-feedforward class

We study global stabilization of strict-feedforward systems with arbitrarily long input delay. These systems may be open-loop unstable but cannot exhibit finite escape instability, providing for a possibility of global stabilization even in the presence of long delay. We derive predictor-based feedback laws for exact compensation of input delay. These feedbacks are given explicitly due to the fact that a predictor state can be written explicitly for strict-feedforward systems. Stability is established in the sense of an L₂ norm on the actuator state, which is not necessarily possible for the broader class of forward complete systems which are stabilizable in the presence of arbitrarily long delay in L_¿ norm. The paper includes a third-order worked example which is not feedback linearizable.