Prof. Ying Tan，University of Melbourne

Extremum seeking control(ESC)is a class of data-driven optimization algorithms which can drive the output of an unknown dynamic system to its optimal value with respect to some cost function by using input and output measurements. This talk summarizes our recent work on how to use some model information to improve the performance of the existing ESC algorithms. In particular, a few fast Esc and model-guided ESC algorithms will be presented, followed by their applications in human- prosthetic interfaces and automotive diesel engine phasing control.