Master's Thesis Tobias Beckhölter
Development of a Prototype for Model Predictive Control of district heating gridsCopyright: EBC
The investigation of city districts, espacially with buildings supplied by a district heating grid, is getting more and more import wehen it comes to the sustainable heat supply of buildings. To meet future requirements of an ever-growing efficiency, new control algorithms are needed which ensure the ideal operation of heating grids while considering heat losses and pump capacity.
The aim of this thesis is to develop a prototype based on the approach of Model Predictive Control (MPC). On the basis of existing and, in part, extensions of building as well as grid models, a dynamic simulation model in the modelling language Modelica, which uses the open-source library AixLib, emerges. A genetic algorithm optimizes the operational parameters of the heating grid. This optimization determines the control variable for the supply system’s maximum efficiency referring to primary energy consumption and costs. Thus, a Model Predictive Controls emerges which is tested by means of a Use-Case, based on the data of a small local heat network in Ergolding, Bavaria. By using the python-based platform JModelica, the suitability of the applied models for numerical optimization is also evaluated.