Artificial neural network modeling of a photovoltaic-thermal evaporator of solar assisted heat pumps

In this work, the artificial neural network model was developed to predict the energy performance of a photovoltaic-thermal evaporator used in solar assisted heat pumps. The experiments were carried out under the meteorological conditions of Coimbatore city (latitude of 10.98°N and longitude of 76.96°E) in India. The energy performance parameters of a photovoltaic-thermal evaporator such as, evaporator heat gain, solar energy input ratio, photovoltaic efficiency and photovoltaic panel temperature were observed with reference to four ambient parameters such as, solar intensity, ambient temperature, ambient wind velocity and ambient relative humidity. The experimental results were used as training data for the network. The multilayer feed forward network is optimized to 4-15-4 configuration for predicting the energy performance of the photovoltaic-thermal evaporator. Analysis of variance was carried out to identify the significant ambient parameter influencing the energy performance of photovoltaic-thermal evaporators. The network predictions are found to be closer to the experimental values with the maximum fraction of absolute variance values, minimum root mean square errors and minimum coefficient of variance values. The analysis of variance results confirmed that solar intensity and ambient temperature are the most influencing parameters affecting the energy performance of photovoltaic-thermal evaporators.