FEIT Research Project Database

Real-time adaptive optimisation

Project Leader: Jonathan Manton
Primary Contact: Jonathan Manton (jmanton@unimelb.edu.au)
Keywords: optimisation; signal processing
Disciplines: Electrical & Electronic Engineering
Research Centre: Nonlinear Signal Processing Lab

This project focuses primarily on the following generic problem. Assume there is a function f that is difficult to invert. By taking a real-time optimisation approach, an inverse g to f can be constructed. However, if f represents the behaviour of a physical device, such as a smart antenna, then f can change over time. This change could be slow (eg, due to temperature) or abrupt (eg, due to mechanical damage sustained in an accident). How can the real-time algorithm adapt to such changes, assuming it knows how f changed? Going further, can the change in f be detected automatically by the algorithm?