Tools for the Control of Modern Solar-Thermal Heating Plants

Among renewable sources, hopes are placed upon solar energy as a viable alternative for sustainable development. Apart from being clean, this source must be profitable and affordable, which can be achieved only if adequate tools are available, in such way that their dynamics can be predicted/analyzed and benefits are maximized. Thus, this short work proposes design tools for modern solar-thermal plants. Firstly, a phenomenological model is drawn, whose parameters are adjusted for a real testbed. Then, considering real data, three different identification procedures are performed (ranging from nonlinear least-squares to frequency- wise methods). Also, in the aims of feedforward compensation applications, an estimation layer is proposed, based on Neural Networks, to predict solar irrandiance disturbances. Simulation results assess that the models are useful for control purposes.