|
|||||
|
|
| 中国锡金属消费量预测方法及应用 | |
| 引用本文： | 刘超,陈甲斌,唐宇,张艳飞.中国锡金属消费量预测方法及应用[J].资源科学,2015,37(5):1038-1046. |
| 作者姓名： | 刘超 陈甲斌 唐宇 张艳飞 |
| 作者单位： | 1. 中国国土资源经济研究院,北京 101149 2. 国土资源部资源环境承载力评价重点实验室,北京 101149 3. 中国地质科学院矿产资源研究所,北京 100037 |
| 基金项目： | 国土资源部地质调查项目:"中国战略性矿产安全评价与支持系统建设"(12120114052901),"矿产资源勘查开发格局及对策研究"(12120114093501),"镍锡钨钼锑对2020年、2025年和2030年国民经济建设保障程度论证与评价"(12120114025301);国土资源部资源环境承载力评价重点实验室开放课题:"榆林国家级能源化工基地资源环境承载力研究"(CCA2015.10)。 |
| 摘 要： | 基于详实的历史数据和合理的预测模型,科学预测中国锡金属消费趋势,对于国家锡资源管理政策的制定与提升国家资源保障能力具有毋庸置疑的意义。在充分考虑影响锡金属消费的宏观经济环境、中观产业政策以及微观消费市场的基础上,采用灰色关联度分析模型,选取了GDP、空调产量、罐头产量、汽车产量和彩色电视机产量等5个关联度>75%的线性因子支撑BP神经网络预测。BP神经网络模型测算得出2002-2013年我国锡消费量的相对误差最大为10.78%,相对误差绝对值平均数为3.33%,对中长期而言精度较高。预测结果显示参考情景下到2020年、2025年及2030年中国锡金属消费需求量分别为26.59万t、29.63万t及31.65万t。 |
| 关 键 词： | 锡金属 消费预测 灰色关联度 BP神经网络 中国 |
| 收稿时间： | 2015-01-10 |
| 修稿时间： | 2015-02-15 |
Tin metal consumption prediction methods and application |
|
| LIU Chao , CHEN Jiabin , TANG Yu , ZHANG Yanfei.Tin metal consumption prediction methods and application[J].Resources Science,2015,37(5):1038-1046. | |
| Authors: | LIU Chao CHEN Jiabin TANG Yu ZHANG Yanfei |
| Institution: | 1. Chinese Academy of Land & Resource Economics,Beijing 101149,China 2. Key Laboratory of Carrying Capacity Assessment for Resource and Environment,Ministry of Land & Resources,Beijing 101149,China 3. Institute of Mineral Resources,Chinese Academy Geological Sciences,Beijing 100037,China |
| Abstract: | A reasonable prediction of tin consumption based on detailed historical data and reasonable modeling is essential to tin resource management policy settings and tin resource guarantee levels. The macroeconomic environment, industry policy and micro consumption market more or less affect tin metal consumption. Here we selected eight factors: gross domestic product, investment in fixed assets, population, tin production, air conditioning yield, canned food production, auto production, and the production of colour television sets. Among these, the relevance of air conditioning yield, canned food production, auto production, the production of colour television sets and gross domestic product to tin production are all above seventy-five percent. We carefully selected these five factors as essential input variables for BP neural network prediction based on Grey Correlation Theory. The results show that the maximum relative error for tin consumption prediction (data used is from 2002 to 2013) is 10.78%, and the average absolute value of relative error for tin consumption prediction is 3.33% using the BP neural network prediction method. This indicates that the BP neural network prediction model is relatively accurate for the medium and long term prediction of tin consumption. The prediction results show that tin metal demand in China will be 265 900 tons, 296 300 tons and 316 500 tons by 2020, 2025 and 2030 under the reference scenario, respectively. |
| Keywords: | tin metal consumption prediction grey correlation theory BP neural network China |
| 本文献已被 CNKI 等数据库收录！ | |
| 点击此处可从《资源科学》浏览原始摘要信息 | |
| 点击此处可从《资源科学》下载免费的PDF全文 | |