Short-term solar power forecasting using similar condition approach based on euclidean distance with weighted coefficients

Abstract

As a renewable source, solar power presents great potential for applications in power systems. However, due to the randomness of the weather condition, solar power output cannot keep steady. Accordingly when a solar photovoltaic (PV) unit is connected to the grid, it will cause frequency deviation in the power system and if the injection level is high, it can be a serious problem in terms of power quality. Hence it is necessary and important to find a way to make short-term forecasting for solar power.

This project aims to make short-term solar power forecasting and contains three main objectives. Firstly, establish a database for real-time solar power and weather data. Secondly, analyze the relationships between each weather factor and solar power output. Lastly, forecast the trend of solar power based on historical data and the similar condition approach.

This report begins with a literature review, introducing some basic theories on solar power forecasting and management of microgrid, followed by the introduction of the similar condition approach and Euclidean distance. The brief introduction of the software used in this project is also included.

Data analysis provides a clear indication of the relationships between each weather factor and solar power output. A case study is provided in this report. Areas recommended for future works are also included.