消谐PWM模型中雅柯比矩阵LU分解的FPGA实现

LU Decomposition of Jacobi Array for Harmonic Elimination PWM Model by FPGA

在逆变电源消谐脉宽调制 (PWM )模型的求解中 ,雅柯比矩阵的求逆是影响求解速度的主要因素 .为加速矩阵求逆 ,本研究在简要分析Gauss消去法的基础上 ,构造出一种基于二维正方心动阵列的矩阵LU分解的电路结构 .通过硬件描述语言对其建模与仿真 ,并经可编程门阵列芯片实验验证 ,结果表明这种结构能快速实现矩阵LU分解 ,从而有效地提高了消谐模型数学求解的实时性 .

In the solution of the harmonic elimination PWM mo del for inverted power supply,the inverse calculation of Jacobi matrix is the main factor affecting the solving speed.In this paper,to accelerate the solving speed of Jacobi matrix,the Gausssian elimination was briefly analyzed,on the basis of which a circuit structure of LU decomposition realized by a two-dimension systolic array was constructed.The structure was then modeled and simulated by VHDL language and its validity was verified by FPGA experiments.The results show that the proposed structure can be used to realize the rapid LU decomposition of the matrix,thus effectively improving the real-time property of the solution of the harmonic elimination model.