Using History Based Probabilistic Irradiance Forecasts for Supporting the Predictive Control of Solar Thermal Systems
Abstract
Results from a modeling study, aimed to investigate into the possibilities for including a model predictive control MPC scheme in a simply structured combisystem are reported. The system modelled uses a flat plate collector system, a single storage - assumed to be well mixed - and an electrical auxiliary energy. The control can act on the aux. heating for the storage tank. A fixed set point thermostatic control gives the benchmark for the MPC control applied. As MPC asks for predictions of the disturbance of the system – which here are the variations of the meteorological operation conditions of the collector system and the heating load governed by the ambient temperatures (and general the load requirements) – two forecast options are applied here, the hypothetical perfect knowledge of the future conditions and a simple forecast scheme, based on a statistical analysis of historical data of the location of interest. As quality measures for the control schemes, the aux. energy consumption and the deviation of the room temperature from a desired level are used to assess the different control schemes.
Keywords
Solar combisystem, Model predictive control (mpc). irradiance forecast