Nonlinear constrained predictive control applied to a coupled-tanks apparatus

The paper deals with the development and application to experimental equipment of fast constrained predictive control algorithms based on feedback linearisation. Rather than using the quadratic programming optimisation, a computationally much cheaper alternative based on interpolation and linear programming is deployed. Despite its undemanding computational nature, this algorithm is found to perform well both in simulation, and when applied to an actual couple-tank rig. The advantages of the algorithm are further illustrated by comparison with PID control