Identification of time varying systems

An algorithm that uses an optimal control theoretic approach to identifying time varying systems is presented. The hybrid cost function used by this algorithm removes many of the limitation associated with identification methods based on the familiar error based cost function. This results in improved parameter tracking of a wide variety of systems. Of particular interest is identification of systems and their adaptation mechanisms observed in man-machine applications.