Prediksi Radiasi Matahari Menggunakan Algoritma Deep Gaussian Process

Abstract

Accuracy in solar radiation prediction plays a crucial role in the development of renewable energy and climate research. This study implements the Deep Gaussian Process (DGP) method for solar radiation prediction with forecast horizons of 7, 14, 21, and 28 days ahead in Medan City, Indonesia. In contrast to conventional prediction methods, DGP offers the capability to capture complex non-linear relationships within meteorological data and provides uncertainty estimation, which is crucial for practical applications such as the management of solar power systems. This research utilizes the dataset 'Development of software for estimating clear sky solar radiation in Indonesia', where the implementation models data dynamics and effectively quantifies prediction uncertainty across various forecast horizons. The novelty of this research lies in the application of DGP, the focus on data from the tropical region of Indonesia, the implementation of the model using various training data subsets for each forecast horizon, and the emphasis on uncertainty estimation, which provides a significant contribution to the field of solar radiation prediction.