说明：  PCA（主成分分析）算法被广泛应用于工程和科学研究中，本报告主要从PCA的基本结构和基本原理对其进行研究，常规的PCA算法主要采用线性算法，通过研究论证发现线性的PCA算法存在着许多不足，比如线性PCA算法不能从线性组合中把独立信号成分分离出来，主分量只由数据的二阶统计量—自相关阵确定，这种二阶统计量只能描述平稳的高斯分布等，因此必须对其进行改进，经改进后的PCA算法有非线性PCA算法、鲁棒算法等。我们通过PCA算法在直线（平面）中拟和的例子说明了PCA在工程中的应用。本例子采用的是成分分析中的次成分（方差最小的成分），通过对结果的分析，我们可以看出，利用PCA算法可以得到较好的拟和结果。
(PCA (Principal Component Analysis) algorithm has been widely used in engineering and science research, This report mainly from the PCA and the basic structure of the basic tenets of its research, Conventional PCA algorithm used mainly linear algorithm, found through research and demonstration linear PCA algorithm, there are many inadequate, For example, not linear PCA algorithm from the linear combination of the independent signal components separated, PCA data only from the second-order statistics-auto-correlation matrix to determine, Such second-order statistics can only describe a smooth Gaussian distribution, it is necessary to improve it. After the improvement of the PCA algorithm is nonlinear PCA algorithm, robust algorithm. PCA algorithm we passed the line (plane), and to be example)