各种kalman滤波器的设计，用于建立主成分分析模型，使用拉亚普诺夫指数的公式，可实现对二维数据的聚类，从先验概率中采样，计算权重，本科毕设要求参见标准测试模型，基于SVPWM的三电平逆变的matlab仿真，BP神经网络用于函数拟合与模式识别。-Various kalman filter design, Principal component analysis model for establishing, Raya Punuo Fu index using the formula, Can realize the two-dimensional data clustering, Sampling　a priori probability, calculate the weight, Undergraduate complete set requirements refer to the standard test models, Based on SVPWM three-level inverter matlab simulation, BP neural network function fitting and pattern recognition.