SIR方法在小型二次电池市场分析上的应用

在新型电池的研究与发展过程中 ,日本小型电池市场发展最具代表性。本文采用分片逆回归方法 (SlicedInverseRegression ,简称SIR) ,在非线性模型的假设下 ,利用日本小型二次电池市场的数据 ,对影响小型二次电池市场销售额的主要因素进行综合与提炼 ,以较高的精度将高维自变量集合综合成一维变量 ,并建立了镍镉电池、镍氢电池和锂离子电池市场年销售额的预测模型。由于经SIR降维后 ,三种电池的预测模型均为一元回归模型 ,因此更容易观察和比较三种电池市场的发展趋势

The application of SIR method in the small-size battery market

In the study of new fashioned batteries, the small size battery market in Japan represents the developing trend of the small size battery market in the world. Under the hypothesis of nonlinear models, this paper synthesizes seven major variables which mainly affect the small size secondary battery sales volumes in Japan and applies the Sliced Inverse Regression (SIR) method to reduce the seven dimensional variable set into a one dimensional variable in fairly high accuracy. After that three predicting models are obtained on yearly sales volumes for Cd Ni battery、MH Ni battery and LiB battery. These three models are all simple regression models, so it is much easier to observe and compare the developing trends of these three kinds of batteries.