Affiliation of Author(s):计算机科学与技术学院/人工智能学院/软件学院

Journal:INTELLIGENT DATA ENGINEERING AND AUTOMATED LEARNING - IDEAL 2017

Key Words:Interval-valued load forecasting Body amenity indicator Multi-output LSSVM SOA Fresh degree function

Abstract:Accurate forecasting of mid-term electricity load is an important issue for risk management when making power system planning and operational decisions. In this study we have proposed an interval-valued load forecasting model called SOA-FD-MLSSVM. The proposed model consists of three components, the Human Body Amenity(HBA) indicator is introduced as the input of meteorological factors, Fresh Degree(FD) function is brought into the forecast method based on setting different weight on the historical days and Least Squares Support Vector Machine based on Multi-Output model, called MLSSVM, to make simultaneous interval-valued forecasts. Moreover, the MLSSVM parameters are optimized by a novel seeker optimization algorithm(SOA). Simulations carried out on the electricity markets data from Jiangsu province. Analytical results show that the novel optimized prediction model is superior to others listed algorithms in predicting interval-valued loads with lower U-I, ARV(I) and MAPE.

ISSN No.:0302-9743

Translation or Not:no

Date of Publication:2017-01-01

Co-author:Zheng, Huiting,Zhao, Chang

Correspondence Author:Yuan Jiabing