School of Computing, Engineering and Technology

Data-Driven Control of Complex Networks and Networked Systems

This project will develop data-driven tools and techniques to control complex networks and interconnected systems where precise modelling of the dynamics, and their interconnection becomes challenging. This problem will be studied in the context of robotics, autonomous and multi-agent systems, power distribution and other complex networks.

Award

MRes | MPhil | PhD | MSc by Research

Start Date

Usually February and October - at individual School's discretion

Research Opportunities

Summary

The focus of this research is on data-driven control of complex networks and networked systems. Networks and multiagent systems have been researched in the context of formation control, consensus algorithms, coordination etc which have various application for robotic and autonomous systems, power grid, sensor and communication networks etc. The ability to manipulate and regulate complex network of interconnected systems requires efficient and robust control algorithms, understanding and availability of tractable model of the systems and network dynamics. Modelling of interaction between potentially unknown heterogeneous systems in a network is a challenge and developing accurate models of large networks through first principles or through system identification will often have modelling errors leading to poor control solutions. In this research direct or indirect data-driven control techniques will be explored fundamentally and in an applied approach to develop optimal solutions for control of complex interconnected and networked systems. Applications of the proposed tools and techniques could be studied through various problems such as cooperative control of robotic swarms of aerial, ground and manipulators systems, optimal power flow in power grids, control of homogeneous and heterogeneous multi-agent systems including sensor and communication networks.