基于实值MUSIC算法的电力谐波分析方法

现代谱估计方法逐渐被应用于电力谐波及间谐波的超分辨率检测与分析,其中的多重信号分类方法(MUSIC)算法最具代表性.然而,常规谱MUSIC算法在分析电力谐波时必须将实值信号转换为复频率信号,因而计算复杂度较高;此外,其伪频谱的峰值搜索过程也存在着类似栅栏效应,导致其频率分析精度有限.为了改善谱MUSIC算法对电力谐波的分析精度,本文基于子空间分析理论提出了针对实值周期信号的谱MUSIC频率检测方法,推导并给出了其伪谱函数的表达式.在此基础上,还利用Newton-Raphson算法对伪谱峰值进行迭代求精,进一步提高了谐波频率的估计精度.仿真结果表明:实值MUSIC伪谱更能有效区分真实谐波谱峰与噪声伪峰,同时本文方法的计算复杂度远远低于原有复值谱MUSIC算法,而频率估计精度与求根MUSIC算法接近,非常有利于电力谐波参数的准确获取.

Power Harmonic Analysis Based on Real-Valued Spectral MUSIC Algorithm

In recent years,modern spectral estimation methods,especially the MUSIC(Multiple Signal Classification) algorithms,are gradually used for high-resolution power harmonic analysis.However,most of them are proposed to detect frequencies of complex signals so that any real-valued signal should be transformed into complex form.The preprocessing may lead to higher computation burden.In addition to this,the picket-fence effect also exists in searching spectrum peaks,which causes low analyzing precision.To overcome...