Optimization of PV array inclination in India using ANN estimator: Method comparison study

Abstract

The network of solar power plants in distributed power system (in India) is growing exponentially. Since tilt angle affects the energy collection by solar panels, the conduction of experiments to find optimum tilt angle is expensive and time consuming in each location. Although different non-linear, empirical models have been proposed by different researchers in India, they have too many constraints and needs complex and rigorous computational effort each time. In this paper, a data driven approach such as ANN estimator has been proposed to estimate (βopt) at which solar energy collection of Photo-Voltaic Array (PVA) will be maximum at selected cities of India. The latitude (ϕ) elevation (EL) from sea level and monthly mean hourly and daily global solar irradiance (Hg) of location are taken into account for this study. The ANN estimator takes Hg, ϕ and EL of site as inputs which are easily available and estimate optimum tilt angle almost instantaneously while testing. The proposed ANN estimator results are compared with analytical results. To validate the performance of ANN estimator results, statistical analysis study is carried out. Where, two statistical approaches such as descriptive method and Altman–Bland methods are applied. The ANA estimator results are found much close to ANA results at 95% confidence interval under statistical study.