There is significant motivation on both economic and environmental grounds to optimise the efficiency of various engineering systems. Extremum-seeking (ES) is a method for achieving real-time optimisation of a system's performance even when the model of the underlying system is not available. Past, present and future ES applications considered at the University of Melbourne include thermoacoustic noise reduction in gas turbine combustors, skin friction drag reduction, optimisation of flex-fuel engines and traffic light scheduling. This seminar will introduce the typical features of ES as well as highlight some of the limitations of traditional ES approaches. New developments that address these limitations will be the focus of the presentation. These developments include "Newton-like" ES, which allows the practitioner to more reliably assign a convergence speed to the closed-loop, and "fast" ES which allows a considerable increase in the convergence speed of the closed-loop.
Dr Will Moase is currently a Postdoctoral Fellow in the Department of Mechanical Engineering at The University of Melbourne. He has previously completed a B.Eng (Hons) and B.CS in 2004, and a PhD in 2009, all at The University of Melbourne.
Dr Moase is involved with research on high dimensional extremum-seeking and its application to skin friction drag reduction in high Reynolds number flows.
Research interests include:
extremum-seeking
flow control
thermoacoustic modelling