Nonlinear Model Predictive Control of a thermo-hydraulic plant: design and experiments

This thesis concerns the modelling and controlof an energy system, i.e. District Heating System(DHSs). In fact, with the European Uniontargets to reduce the use of fossil fuels by 2050,interest in DHSs has increased [3]. DHSs shouldcover at least 50% of the heating demand ofmost European countries by 2050, achieving additional30% of energy savings in the heatingsector with respect to 2015 [4].District heating means a system of distanceheating units that, through a network of waterpipelines, transports the heat generated by largethermal plants to individual residential units fordomestic heating. District heating reduces energywaste and has a lower environmental impactthan traditional heating methods. Thus,it is clear how optimization of these systems isnecessary to have better energy efficiency.The case study investigated in this thesis isa scaled-down thermo-hydraulic experimentalplant designed and built with the contributionof the National Research Council (CNR) whichit can be regarded as a reproduction of a DHSFigure 1.The aim of this work is to design a Model PredictiveControl (MPC) algorithm to optimizethe behaviour of the system. The peculiarityof this type of control lies in the fact thatcontrol actions are obtained as a result of anoptimization problem and, thus, the possibilityof introducing constraints on inputs, outputsand states.In particular, the aim is to track temperaturereferences in the plant, by optimizing theperformance of control variables.