基于VMD-LSTM的舟山港低硫保税船用燃油价格预测  

Forecast of Low-Sulfur Marine Fuel Price in Zhoushan Port Based on VMD-LSTM

作　　者：陈亮 沈良朵 陆凡 班文超 CHEN Liang;SHEN Liangduo;LU Fan(School of Naval Architecture and Maritime of Zhejiang Ocean University,Zhoushan 316022;School of Marine Engineering Equipment of Zhejiang Ocean University,Zhoushan 316022,China)

机构地区：[1]浙江海洋大学船舶与海运学院,浙江舟山316022 [2]浙江海洋大学海洋工程装备学院,浙江舟山316022

出　　处：《浙江海洋大学学报（自然科学版）》2023年第4期346-351,共6页Journal of Zhejiang Ocean University：Natural Science

基　　金：国家自然科学基金(52101330);舟山市科技局一般项目(2023C41017)。

摘　　要：低硫船用燃油在航运成本中的高占比,其价格走势关系到众多船舶相关方的利益,关注价格走势有利于相关方控制航运成本。本文以舟山地区的低硫保税船用燃油价格数据为基础,对几种预测模型结果进行比对,发现VMD-LSTM模型性能最佳,可以为船用燃油期货交易提供决策思路。Low sulfur marine fuel oil has high proportion in the shipping cost,and its price trend is related to the interests of many shipping related parties.Thus it is conducive for the relevant parties to control the shipping cost by paying attention to the price trend.Based on the price data of low-sulfur marine fuel in Zhoushan area,this paper compares the results of several prediction models,and finds that the VMD-LSTM model has the best performance,which can provide better decision-making ideas for marine fuel oil trading.

关 键 词：VMD-LSTM 变分模态分解 长短期记忆神经网络 船用燃油价格 

分 类 号：U677.2[交通运输工程—船舶及航道工程] F416.22[交通运输工程—船舶与海洋工程] F764.1[经济管理—产业经济]