GAMMA PROBABILISTIC NEURAL NETWORK MODELS TO TRACK OPTIMAL THERMAL AND ELECTRICAL POWER IN SOLAR PVT SYSTEMS

Abstract

Solar PVT systems combine the characteristics of the photovoltaic and thermal solar systems in a single module. Due to the abundant presence of the natural resource from the sun–solar energy, in the past decade several algorithms and related electronic approaches were developed in order to monitor the photovoltaic and thermal panels maximum power generation.  Solar PVT Systems possess several control parameters designed to produce better results and in this paper, the task is to track the optimal thermal and electrical power. As such, no appropriate control mechanism has been developed for tracking the maximum power generated from Solar PVT systems.  In  this  paper,  a  PVT  control  algorithm  based  on  the  proposed  neural  network architectures are designed to compute  the  Optimal  Power  Operational  Point  (OPOP)  by  taking  into  account  the  model behavior of the Solar PVT system. Ambient temperature and irradiation are considered by the optimal power operational point to compute the optimal mass flow rate of Solar PVT module. Numerical simulation results prove the effectiveness of the proposed neural network models compared with that of the calculated outputs and the solutions derived from the earlier literature studies.