摘要
从我国CO2排放量的不确定性、不完整性、小样本等特征出发,以灰色系统(GM)模型和支持向量机(SVM)模型为基础,建立基于粗糙集的组合预测模型.利用该模型以我国1990~2011年CO2排放量的数据以及同期的人口数量、GDP和能源消耗总量数据为基础对我国同期CO2排放量进行预测来验证其有效性,最后对我国2012~2017的CO2排放量进行预测.结果表明,灰色系统理论与支持向量机模型仅能够反应我国CO2排放的长期变化趋势,在预测精度上存在一定缺陷而基于粗糙集与灰色SVM的组合预测模型在预测精度上明显优于以上两种方法,能够对我国未来CO2排放量进行准确有效的预测分析.
The text is to establish a combined forecasting model based on the Rough Set, the Grey System (GM) Model and Support Vector Machine (SVM), in consideration of the Chinese CO2 emission:uncertainty, imperfection and small sample properties. With this model, we predict the same period Chinese CO2 emissions to verify the effectiveness of combination forecasting model, based on the data of the Chinese CO2 emissions from 1992 to 2011as well as the population, GDP and total energy consumption in the same period, then predict our country future CO2 emissions from 2012 to 2017. The results show that GM and SVM are only able to show long-term trend of CO2 emissions, combination prediction model in forecasting accuracy is better than the above two methods, this method can effectively analyze Chinese future CO2 emissions.
出处
《中国环境科学》
EI
CAS
CSCD
北大核心
2013年第12期2157-2163,共7页
China Environmental Science
基金
国家自然科学基金资助项目(71071052)
2012年河北省社会科学基金课题(HB12YJ014)