NON-CONVEX OPTIMISATION VIA EXTREMUM SEEKING

The scientific community has recently renewed its interest on Extremum Seeking, which is an optimisation technique that has been investigated since the 1920s.

This optimisation technique has been largely used for solving industrial problems. Indeed, one of its key features is that the optimisation is made on-line and without the knowledge of the cost’s gradient.

Especially useful for physical plants for which a digital twin is not available, Extremum Seeking relies on the concepts of averaging and time-scale separation between exploration and exploitation. In this talk, I will review the basics of Extremum Seeking and show how recent advancements enable its application to non-convex optimization problems.

Speaker:

Nicola Mimmo (Member, IEEE) is an Assistant Professor at the Department of Electrical, Electronic, and Information Engineering, University of Bologna. He teaches "Automatic Control" and "Ground Vehicle Dynamics" in master's degree programs, and "Optimisation via Extremum Seeking" and "Kalman-like Filtering" for PhD national programs. From 2011 he has been working on national and European projects to develop unmanned aerial vehicles for civil applications. In 2006 and 2014, he collaborated with major national and European aircraft companies to develop flight control laws currently protected by international patents. He is currently involved in the automatisation of test facilities of automotive companies. He promoted and is responsible for international exchange programs with the University Claude Bernard Lyon 1, France. He organised invited sessions at the IFAC Symposium on Automatic Control in Aerospace, and participated in the organisation committee of the IFAC Workshop on Aerospace Control Education. He is an associate editor for the IEEE-CSS International Conference on System Theory, Control and Computing. He participated in the organisation of the European Robotic Forum ERF2024. His research interests range from fault-tolerant controls to non-linear control systems in aerospace and automotive. Recently, he also focused on optimisation via extremum seeking and distributed optimisation.

26.02.2025