基于 LSTM 的上海市天然气需求量预测

打开文本图片集

引用格式：，等. 基于 LSTM 的上海市天然气需求量预测[J]. 当代化工，2026, 55（4）：942-946.

LI Kaihua, WANG Jinpeng, ZHANG Jie, et al. Prediction of Natural Gas Demand in Shanghai Based on LSTM[J]. Contemporary Chemical Industry, 2026, 55(4): 942-946.

1 ， 2 ， 1 ，秦川 2 ，高福特 1 ，李丹 1 ，张博钧 1

（1. 300450；

2. 102299）

中图分类号：TQ-9；TE82

文献标志码：A

文章编号：1671-0460（2026）04-0942-05

Prediction of Natural Gas Demand in Shanghai Based on LSTM

LI Kaihua 1, WANG Jinpeng 2, ZHANG Jie 1, QIN Chuan 2, GAO Fute 1, LI Dan 1, ZHANG Bojun 1

PipeChina Engineering Technology Innovation Co., Ltd., Tianjin 300450, China;

2. College of Mechanical and Transportation Engineering, China University of Petroleum (Beijing), Beijing 102299, China)

Abstract: Clean and low-carbon natural gas plays a significant role in China's energy transition, its demand continuously increases, accurate and rational forecasting of natural gas demand helps ensure supply-demand balance, infrastructure planning, and policy formulation. In this paper, key influencing factors of natural gas demand in Shanghai were identified through grey relational analysis and the Pearson correlation coefficient method. An LSTM model was subsequently constructed to predict natural gas demand. The results indicated that ten factors including the added value of the tertiary industry, regional GDP, per capita disposable income of urban residents, and urban natural gas pipeline length were critically influential. It was projected that Shanghai's natural gas demand would maintain sustained growth in the coming years, reaching 13.514 billion m 3 in 2025.

Keywords: Natural gas; Prediction; LSTM; Gray correlation; Influencing factors; Gas; Simulation; Sustainability

天然气作为高效、清洁的绿色能源，正日益成为全球能源转型的重要组成部分[1-3]。（剩余10165字）