Estimation of Hourly Solar Radiations on Horizontal Surface from Daily Average Solar Radiations Using Artificial Neural Network

Hourly solar radiation information is necessary to increase the effectiveness of photovoltaic (PV) energy systems. However, compared with daily average radiation, measurement hourly much less available. In this study, a multilayer perceptron artificial neural network based on proposed estimate amount radiation. The method also relies location PV system, hour when needed, and month needed. Two separate networks were built direct diffuse summation which estimated value global trained using three years data from 2016 2018 five locations around Japan. taken website Japan Meteorological Agency (JMA). number hidden neurons for each was determined by comparing regression obtained during training process. validated actual measured in 2019 same used networks. reliability confirmed minimum R2 0.951 estimation 0.983 Further improvement accuracy estimating peak considered improve total model future.