中国人口资源与环境
中國人口資源與環境
중국인구자원여배경
China Polulation.Resources and Environment
2013年
10期
127~133
,共null页
煤炭市场 整体经验模态分解 支持向量机 价格预测
煤炭市場 整體經驗模態分解 支持嚮量機 價格預測
매탄시장 정체경험모태분해 지지향량궤 개격예측
coal market; Ensemble Empirical Mode Decomposition (EEMD) ; Support Vector Machine (SVM) ; price forecasting
近年来我国己成为世界第一煤炭生产国、消费国和进口国.2013年实施电煤价格并轨后,我国国内煤价与国际煤价的波动趋势也将日趋一致.在国内煤炭供给过剩的情况下,我国煤炭进口却仍在大幅增加,对我国煤炭市场造成了较大影响.因此,准确预测国际煤炭价格既有利于择机进口煤炭,也有助于进一步完善国家煤炭应急储备机制.通过整体经验模态分解(EEMD),非平稳煤价序列被分解重组为正常市场波动、重大事件影响、长期趋势等具有不同经济含义的时间序列.这三项时间序列被用于支持向量机(SVM)组合建模对国际煤价进行了一年内短期预测.研究发现:2013年国际煤炭价格将呈整体下跌趋势,趋向长期趋势线,并处于低位小幅波动.
近年來我國己成為世界第一煤炭生產國、消費國和進口國.2013年實施電煤價格併軌後,我國國內煤價與國際煤價的波動趨勢也將日趨一緻.在國內煤炭供給過剩的情況下,我國煤炭進口卻仍在大幅增加,對我國煤炭市場造成瞭較大影響.因此,準確預測國際煤炭價格既有利于擇機進口煤炭,也有助于進一步完善國傢煤炭應急儲備機製.通過整體經驗模態分解(EEMD),非平穩煤價序列被分解重組為正常市場波動、重大事件影響、長期趨勢等具有不同經濟含義的時間序列.這三項時間序列被用于支持嚮量機(SVM)組閤建模對國際煤價進行瞭一年內短期預測.研究髮現:2013年國際煤炭價格將呈整體下跌趨勢,趨嚮長期趨勢線,併處于低位小幅波動.
근년래아국기성위세계제일매탄생산국、소비국화진구국.2013년실시전매개격병궤후,아국국내매개여국제매개적파동추세야장일추일치.재국내매탄공급과잉적정황하,아국매탄진구각잉재대폭증가,대아국매탄시장조성료교대영향.인차,준학예측국제매탄개격기유리우택궤진구매탄,야유조우진일보완선국가매탄응급저비궤제.통과정체경험모태분해(EEMD),비평은매개서렬피분해중조위정상시장파동、중대사건영향、장기추세등구유불동경제함의적시간서렬.저삼항시간서렬피용우지지향량궤(SVM)조합건모대국제매개진행료일년내단기예측.연구발현:2013년국제매탄개격장정정체하질추세,추향장기추세선,병처우저위소폭파동.
In recent years,China has become the largest coal producing,consuming and importing country in the world.After the mergence of domestic coal prices in 2013,the domestic coal price trend gradually matches with the international one.In spite of the excess supply of the domestic coal,the coal imports are still in a substantial increase,which has a big impact on the domestic coal market.Therefore,to accurately predict the international coal price is not only conducive to choose good opportunities to import coal,but also helpful to prevent further improve the national emergency coal reserve mechanism.The non-smooth series of international coal prices is decomposed and restructured to three items with different economic implications by the Ensemble Empirical Mode Decomposition (EEMD) technology,such as the normal market fluctuations,major events and long-term trend.Then the three items are used to model the combined Support Vector Machine (SVM) to forecast coal price in one year.The study found that the intemational coal price will fall towards the long-term trend line and be in low fluctuations in year 2013.