Forecasting of Solar Resouce and Load Demand Using the Hourly-based Prediction Method

摘要

This paper presents a comparative study of the solar irradiance and load demand forecasting methods for a hybrid energy system. Two forecasting techniques were adopted to predict the hourly solar irradiance and load demand. The first technique is the commonly used Single Exponential Smoothing (SES) forecasting method which computation is based on previous day data. The other technique is the Hourly-Based Prediction (HBP) method. The principle underlying this method is to involve both of the previous and current day data for predicting the new values. As the latter takes the most recent hourly data into account, which is adjacent to the prediction hour, it is able to predict appropriately despite the stochastic nature of the solar irradiation. The SES prediction is used as a benchmark to evaluate the HBP method and the HBP is found to outperform the SES for predicting the solar irradiance and both of the techniques perform well for the load demand forecast using low smoothing constant.