Wind Speed and Solar Irradiance Prediction Using Advanced Neuro-Fuzzy Inference System

摘要

Fluctuations in wind speed and solar irradiance results in power quality issues in grid connected wind mills and solar PV generation systems respectively. Accurate prediction of wind and solar energy requires efficient energy management in smart grid. This paper uses a regularized neuro-fuzzy system for accurate prediction of wind speed and solar irradiance. Proposed method, regularized extreme learning adaptive neuro-fuzzy system (RELANFIS) combines the learning capability of extreme learning machine (ELM) and knowledge representation of fuzzy inference system. The membership function parameters of RELANFIS are randomly assumed in a constrained range and consequent parameters are calculated analytically. The prediction capability of RELANFIS is tested against well-known ELM based neuro-fuzzy system (OS-Fuzzy-ELM) and other kernel based systems.