Interval forecasting of renewable power generation

• Main Author: Luo, Lingfeng
• Other Authors: Xu Yan
• Format: Final Year Project
• Language: English
• Published: 2018
• Subjects: DRNTU::Engineering::Electrical and electronic engineering::Electric power::Production, transmission and distribution
• Online Access: http://hdl.handle.net/10356/74512
• Institution: Nanyang Technological University

With the development of renewable power generation, forecasting of renewable power generation output is significant for modern power grids. Solar power generation, which is the important component of renewable power generation, is selected as the analysis direction in this project. The solar incoming radiation (SIR), the main factor of uncertainty for solar power generation, is used as training and testing data. By comparing the forecast accuracy of gradient descent, normal equation and extreme learning machine (ELM), optimal result of point forecasting is provided, which is the initial data for interval forecasting. The main process is the conversion from value of point forecasting to prediction intervals (PIs) with the help of hourly standard deviation (HSD) of SIR. Because of HSD is connected to the uncertainty of SIR, PIs is capable of reduce the uncertainty of output of solar power generation. This method has good reference value for other renewable power generation with high uncertainty.