Forecasting of Daily Total Horizontal Solar Radiation Using Grey Wolf Optimizer and Multilayer Perceptron Algorithms

Solar radiation data is an indispensable input for photovoltaic and solar-thermal systems. In this regard, the consistent solar radiation forecasting is a primary task in solar energy applications. In this paper, grey wolf optimizer algorithm is integrated to the multilayer perceptron algorithm in order to forecast the daily total horizontal solar radiation. In the forecasting phase, air temperature, relative humidity and diffuse horizontal solar radiation parameters are evaluated in 3-tupled and 2-tupled input structure. In addition, the accuracy of the hybrid forecasting model developed is also tested on the basis of the sigmoid, sinus and hyperbolic tangent activation functions employed in the multilayer perceptron algorithm. The forecasting results show that grey wolf optimizer-based multilayer perceptron model is appropriate to predict the daily total horizontal solar radiation, efficiently.