In order to implement the state feedback control laws being produced by many of the advances in nonlinear control system design, state observes will be necessary. This research will seek to develop observers for deterministic, discrete-time or sampled-data nonlinear systems. The main thrust will be to extend the work on Newton-like observers in order to increase their applicability: an improved region of convergence and a better understanding of how to efficiently perform the computations in a "real-time" setting. An important goal will be to do this in the context of solving an actual engineering problem. In cooperation with engineers from Ford Motor Co., a torque equalization problem has been identified. The initial goal will be to design and test an observer on a simulation model, and then in an engine dynamometer test cell, in order to evaluate the achievable performance of an observer-controller combination.
