Affiliation of Author(s):经济与管理学院
Journal:PROCEEDINGS OF 2017 IEEE INTERNATIONAL CONFERENCE ON GREY SYSTEMS AND INTELLIGENT SERVICES (GSIS)
Key Words:multi-variables grey model TMGM(1,N) driving variables convolution integral total electricity consumption
Abstract:Electricity demand prediction plays an important role in the policy makings and plans for the governments, energy sector investors and other relevant stakeholders. Although there exist several forecasting techniques, selection of the most appropriate technique is of great importance. One of the forecasting techniques which has proved successful in prediction is GM(1, N). In order to clarify the interaction mechanism of driving variables and improve the accuracy of the model, a new model which is based on the development trend of multiple driving variables, abbreviated as TMGM (1, N), is proposed. Firstly, a new forecast model of the development trend of the driving variables is established in order to make better use of the interaction mechanism of the driving variables. On the basis of that, the new grey model TMGM (1, N) is constructed. Meanwhile, the solution to the model parameters are derived on the least square method. And the time response formula is solved by the convolution integral to make up the defects of the solving method of traditional model GM(1, N). Finally, a real application about the forecast of the total electricity consumption in Jiangsu Province is used to demonstrate the feasibility and practicability of the TMGM(1, N) model. The results indicate the superiority of TMGM(1, N) model when compared with GM(1, N) model and TGM(1, N) model.
Translation or Not:no
Date of Publication:2017-01-01
Co-author:dsb,Zhao, Kai
Correspondence Author:dyg
Date of Publication:2017-01-01