摘要
提出一种实数形式的求根多重信号分类(Root-MUSIC)算法,其无需将实值转换为复频率,从而降低了计算复杂度。由Root-MUSIC多项式的根估计出信号频率,同时由这些根采用普罗尼(Prony)法中的最小二乘法估计出信号的幅值和相位。与传统先估计频率、再由其他算法估计幅值和相位的方法相比,所提方法减少了大量计算步骤。仿真结果表明,用所提方法估计的间谐波参数精度高,在存在噪声的情况下也能准确估计出谐波和间谐波频率、幅值和相位。
A Root-MUSIC（Root Multiple Signal Classification） algorithm in real form is proposed,which avoids the transformation of real value into complex frequency to reduce the computational complexity.The signal frequencies are estimated from the roots of Root-MUSIC polynomial,and the amplitudes and phases of signal are estimated at the same time based on the least square method of Prony method.Since the traditional methods estimate the frequencies first and then estimate the amplitudes and phases by other algorithms,the proposed method cuts down the calculation steps greatly.Simulative results show that,the proposed method has high accuracy and can accurately estimate the frequency,amplitude and phase under noisy conditions.
出处
《电力自动化设备》
EI
CSCD
北大核心
2012年第11期56-59,71,共4页
Electric Power Automation Equipment
基金
中央高校基本科研业务费资助项目(CDJZR12150025)~~
