求解最大特征值及其对应特征向量的FPGA实现  

The Implementation of Solving the Maximum Eigenvalue and the Maximum Eigenvector Based on FPGA

作　　者：张雪[1] 金铭[1] 张敏[1] 

机构地区：[1]哈尔滨工业大学(威海),威海264209

出　　处：《科学技术与工程》2012年第14期3500-3502,3519,共4页Science Technology and Engineering

摘　　要：针对某些领域只需求解矩阵的最大特征值及其对应特征向量的特点,设计了基于乘幂法的复矩阵的最大特征值及其对应特征向量求解的FPGA实现,提高了运算速度。设计采用状态机设计方法,将9×9复矩阵的定点数格式转化为浮点数运算,使得到的特征值及特征向量有很高的精度。结果表明,本设计稳定并可实现工程化应用。Because some areas only need the maximum eigenvalue and maximum eigenvector of the matrix,the maximum eigenvalue and maximum eigenvector of the matrix in the platform of FPGA using power algorithm are solved,which increases the operation speed.The design uses the Finite State Mechines and converts the data format from fixed data to float data which makes the eigenvalue and eigenvector having high precision.The conclusion proves that the design is stable and can be used in the project.

关 键 词：最大特征值 乘幂法 状态机 FPGA 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]