Mark Cannon

Mark’s research is about control and optimization of systems with constraints and model or measurement uncertainty. Constrained optimal control problems have a diverse range of applications as a result physical, environmental and economic restrictions on system operations. Feedback control systems that optimize predicted future behaviour can account for constraints explicitly, and this can provide significant improvements in performance and applicability. He is interested in the fundamental properties of constrained control strategies such as feasibility and closed-loop stability, as well as computational issues such as convexity and efficiency of implementation for systems with fast dynamics and stochastic uncertainty.

Robust adaptive NMPC using ellipsoidal tubes Computationally Tractable Robust Nonlinear Model Predictive Control using DC Programming Deep learning adaptive Model Predictive Control of Fed-Batch Cultivations Learning-based Homothetic Tube MPC A successive convexification approach for robust receding horizon control