Medium-Term Horizon Time Photovoltaic Power Generation Prediction for an Island Zone

Abstract

This article presents a Johansen test assessing the predetermined long-term relationships of datasets from a photovoltaic (PV) plant to predict the power output of an island zone. The goal was to use Johansen’s model to predict the PV power generation in the island of Mauritius. In this article, time series using an on-site measurement dataset have been used to design an original prediction model, the Johansen model for PV power output. This model is trained to predict random monthly, weekly, and daily PV power outputs in different seasons and years. The experimental results demonstrate that the Johansen model is a powerful medium-term predicting tool.

Keywords

• predicting
• cointegration
• Johansen