Global solar radiation time series forecasting using different architectures of the multilayer perceptron model

Abstract

In this work, the multilayer perceptron model was used to forecast the time series of global solar radiation for a near future about a week. Different architectures of this model were built through varying its different hyperparameters such as optimizers, activation functions, number of neurons and neuron dropout in which their performance was evaluated using error metrics. It was found that the architectures (60, SGD, Sigmoid), (10, Adam, Relu) and (60, SGD, Sigmoid) presented an R² around 0.877, 0.873 and 0.872, respectively. The architecture with neuron dropout (150, SGD, Sigmoid, 0.2) presented a higher performance among all the architectures evaluated and its R² value was 0.884. Architectures with higher performance are used to predict future values of solar radiation.