Research on liquefied petroleum gas price prediction using variational mode decomposition and a combined model
BAO Ying
LI Duo
CAO Xin
GONG Wenxin
ZENG Yuting
Abstract：As an important fuel and chemical raw material,liquefied petroleum gas(LPG)plays a crucial role in the daily lives of residents and the industrial development.Accurate forecasting of its market price can help better understand the market supply and demand situation,so as to optimize resource allocation,improve production efficiency and market competitiveness.This paper proposes an ensemble prediction model that uses a combination of variational mode decomposition(VMD),autoregressive integrated moving average model(ARIMA),back propagation(BP)neural network,gate recurrent unit(GRU)and long short-term memory(LSTM)to predict the market price of LPG.Firstly,VMD is used to decompose the price time series data into three IMF series.Secondly,for each IMF series,based on historical price data and influencing factor data,ARIMA,BP,GRU,LSTM algorithms are used to make predictions.Then,the prediction results of these algorithms are weighted and integrated to obtain the prediction results of each IMF series.Finally,the prediction results of all IMF decomposition sequences are synthesized to obtain the prediction results of the combined model.Compared with the single optimization prediction model ARIMA,BP,GRU,and LSTM,the evaluation indexes of the related models have been significantly improved.
Keywords：variational mode decompositionvariational modeARIMABPGRULSTMLPG price prediction
Publication Date：2025-03-28
Online Publishing Date：2025-08-15(First online date of this platform, not the publication date of the document)
Pages：11( 6-16 )