构建基于人工智能的成品油价格预测模型  

Build a Prediction Model for Refined Oil Prices Based on Artificial Intelligence

作　　者：张旭莉 赵星原 郭锴 Zhang Xuli;Zhao Xingyuan;Guo Kai(Beijing Longyu Petrochemical Co.,Ltd.,Beijing 100011,China)

机构地区：[1]北京龙禹石油化工有限公司,北京100011

出　　处：《当代石油石化》2025年第2期45-49,54,共6页Petroleum & Petrochemical Today

摘　　要：随着全球经济快速发展和能源结构持续调整,成品油市场展现出日益复杂多变的特征。价格作为油品销售企业利润波动的关键因素,其准确预测对企业经营决策具有重大意义。传统预测方法主要依赖于宏观经济指标、供需关系等定性分析手段,在捕捉市场细节和趋势方面存在局限性。近年来,人工智能技术的快速发展为成品油价格预测提供了新的视角和工具。构建基于人工智能的成品油价格预测模型,运用深度学习技术挖掘和分析历史数据,提升价格预测的准确性,为企业的经营决策提供科学支撑。以传统的向量自回归模型作为基准,对比不同影响因素和预测周期下的预测精度,全面评估2种模型的性能差异。实验结果表明,基于人工智能的深度学习模型在从历史数据中挖掘潜在规律方面表现出显著优势,预测精度更高。该研究成果能为成品油销售企业提供更精确的价格预测工具,助力企业合理决策,提升经营效益。With the rapid development of the global economy and the continuous adjustment of energy structure,the refined oil market has shown increasingly complex and volatile characteristics.As a key factor in the profit fluctuations of oil sales enterprises,accurate price prediction is of great significance for business decision-making.Traditional forecasting methods mainly rely on qualitative analysis methods such as macroeconomic indicators and supply-demand relationships,but have limitations in capturing market details and trends.In recent years,the rapid development of AI technology has provided new perspectives and tools for predicting refined oil prices.This article constructs an refined oil price prediction model based on AI,using deep learning technology to mine and analyze historical data,improve the accuracy of price prediction,and provide scientific support for business decision-making.In the article,the traditional Vector Autoregression(VAR)model is used as a benchmark to compare the prediction accuracy under different influencing factors and prediction cycles,and comprehensively evaluate the performance differences between the two models.Experiments have shown that deep learning models based on AI exhibit significant advantages in mining potential patterns from historical data,with higher prediction accuracy.This research achievement can provide more accurate price prediction tools for refined oil sales enterprises,helping them make reasonable decisions and improve operational efficiency.

关 键 词：人工智能 成品油价格预测 向量自回归模型 深度学习模型 

分 类 号：F42[经济管理—产业经济]