Stochastic data-driven NMPC for partially observable systems using Gaussian processes: a mineral flotation case study

Abstract

A nonlinear model predictive control strategy for froth flotation under partial observability, using a Gaussian process state-space model that predicts both measured and latent variables and propagates their uncertainty into the optimisation. The framework maintained efficient operation despite frequent changes in feed particle size and measurement noise.

Dr Paulina Quintanilla

Assistant Professor in Process Systems Engineering

Assistant Professor in Process Systems Engineering at UCL, leading the LOOPS research group. Research interests include machine learning, optimisation and control for processes and physical systems.