Artificial neural network coupled building-integrated photovoltaic thermal system for indian montane climate

Abstract

Indian Himalayan Region (IHR) encompass approximately 16.2% of India’s geographical area and characterized as a montane climatic zone. The ambient temperature of this region is extremely low, with a minimum average temperature of 2–18 °C. Major settlements in this region are located at an altitude of more than 1000 m above sea level. Therefore, according to The Energy and Resources Institute (TERI), India, there is excessive thermal and electrical energy consumption in these regions pertaining to space heating. Moreover, the current energy sources utilized are not clean sources and hence contribute towards CO2 emissions. A building-integrated photovoltaic thermal (BIPVT) system is modelled to provide the necessary thermal energy alongside electricity for space heating intended for the climatic conditions of Srinagar. The study utilized an application-centric approach in modelling and analyzing the BIPVT system and, hence quantifying the annual outputs for Srinagar, India. Furthermore, the developed mathematical model is used for training an artificial neural network to predict the annual thermal and exergy outputs of the system based on six different model parameters. Three designs, one operational and two application-based parameters are considered as inputs to the network. Moreover, a multi-criteria decision-making method is employed to determine the best set of BIPVT parameters giving optimum annual thermal and exergy outputs along with electrical, thermal and exergy efficiencies of the modelled system. The modelled neural network performed reasonably well against the unseen test dataset in predicting the BIPVT yield with an R2 value >0.97. An optimum annual thermal and exergy gains of 2,25,459 kWh and 67,132.38 kWh, respectively, are quantified for the best set of BIPVT parameters. The system’s optimum average electrical (power conversion), thermal and exergy efficiencies are observed to be 14.87, 56.28, and 17.59%, respectively.