Research HighlightsMulti-parametric/explicit Model-Predictive Control
One key application of multi-parametric programming is model-predictive control (MPC), where a multi-stage constrained optimization problem is formulated based on the system model and restrictions. As this optimization problem is a function of the initial states of the system, in classical MPC this optimization problem is solved online as soon as the nominal values of the states realize themselves. Conversely, multi-parametric programming allows this optimization problem to be solved offline as a function of the states, reducing the online computational effort to a point location and a function evaluation. As the inventors and pioneers of multi-parametric model predictive control, our group continuously develops novel approaches for multi-parametric/explicit MPC, and uses applications such as periodic systems or cogenerations of heat and power to highlight the capabilities of these control strategies, many of which have been embedded in PAROC, an integrated framework and software platform for the design, optimization and model-based control of process systems. |