This paper proposes an adaptive observer for discrete-time MIMO systems with unknown nonlinear dynamics and time-delay. The case that the system output equation has time-delay and nonlinearities is also studied. By using a high-order neural network (HONN), the exact system model, Lipschitz or norm-boundedness assumptions of unknown nonlinearities are not required. Other constraints used in neural-based observer design, i.e., strictly positive real (SPR) or matching conditions, are also removed. Applicability of the presented scheme is verified by simulation.
|Translated title of the contribution||Nonlinear Observer Design for Discrete MIMO Systems with Unknown Time Delay|
|Title of host publication||47th IEEE Conference on Decision and Control, Shanghai, PR China|
|Pages||6137 - 6142|
|Number of pages||6|
|Publication status||Published - 2009|
Bibliographical noteName and Venue of Event: Shanghai
Conference Organiser: IEEE