文献类型：期刊文献

英文题名：Forecasting the LNG Manufacturing Price Index from a Supply and Demand Perspective: The Case of Peoples Republic of China

作者：Pan, Kai[1]; Xie, Xiang[2]; Zhang, Tian[3]; Han, Kejiang[2]; Wang, Mingfei[4]; Huang, Liming[4]; Guo, Jiayu[2]; Zhang, Yuantao[2]; Luo, Haiwei[5]; Lan, Zhixuan[6]

第一作者：Pan, Kai

机构：[1] China University of Petroleum, Beijing, China; [2] China Petroleum Planning and Engineering Institute, Beijing, China; [3] Beijing Union University, School of Management, Beijing, China; [4] PetroChina Hunan Gas Marketing Company, Changsha, China; [5] North China Electric Power University, Baoding, China; [6] PetroChina Hunan Gas Marketing Company, Beijing, China

第一机构：China University of Petroleum, Beijing, China

年份：2025

外文期刊名：International Conference on Electrical, Computer, and Energy Technologies, ICECET 2025

收录：EI(收录号：20261820619226)

语种：英文

外文关键词：Costs - Crude oil price - Decision making - Forecasting - Liquefied natural gas - Natural gas transportation - Natural gasoline plants - Neural networks - Production control

摘要：As a low-carbon fossil energy source, natural gas plays a crucial role in mitigating climate change and improving air quality, making it a key element in China's strategy to achieve its "carbon peaking and carbon neutrality"goals. Liquefied natural gas (LNG) has become a vital component of the global energy system due to its favorable transportation and storage characteristics and wide range of applications, particularly in facilitating the transition of energy structures. With the continuous increase in LNG consumption, accurately forecasting LNG manufacturing prices has become a central issue in the industry, as it provides a crucial reference basis for both enterprise production and industry development. This paper analyses the LNG ex-factory price, identifies six influencing factors including international crude oil futures prices, domestic crude oil spot prices, international natural gas futures prices, CSI 300 Index, domestic LNG production and domestic LNG sales based on typical correlation analysis. Using historical LNG production price data from 2016 to 2022, a Nonlinear Auto Regressive with Exogenous Inputs (NARX) neural network model is developed, integrating both historical price data and the identified influencing factors. The performance of the NARX model is compared to a traditional BP neural network model, demonstrating its high accuracy and stability in forecasting LNG production prices. The findings suggest that the proposed forecasting framework can effectively guide LNG producers in price forecasting and production planning, providing more accurate decision-making support for the industry. ? 2025 IEEE.