摘要
能源消费弹性系数反映了一个国家能源消费增长速度与国民经济增长速度之间的比例关系,是衡量一个国家能源利用效率的重要指标。鉴于数据中存在较多的噪声,首先用小波分析方法对数据进行滤波,然后用滤除了噪声的数据作为输入变量,用支持向量回归方法建模,并预测未来10年我国能源消费弹性系数变化的规律。实际数据检验表明,该预测方法还是可行的。
Elasticity coefficient of energy resource consumption reflects the increasing rate proportional relations between energy resource consumption and national economy development in a country, which is an important index used to measure the utilization effect of energy resource. Considering the ＂noises＂ existed in the accumulated datum, this paper erases the noises by wavelet transform first, then builds the SVR model based on processed datum, and at the end, predicts the elasticity coefficients of energy resource consumption during the nearest 10 years. The demonstration analysis illustrates that the prediction method is completely feasible.
出处
《沈阳航空工业学院学报》
2007年第3期78-80,68,共4页
Journal of Shenyang Institute of Aeronautical Engineering
基金
国家自然科学基金项目资助(70471074)
关键词
支持向量回归
统计学习理论
小波分析
能源消费弹性系数
support vector regression
statistical learning theory
wavelet transform
elasticity coefficient of energy resource consumption
