基于Root-MUSIC和Adaline神经网络的 间谐波参数估计

为了提高电力系统间谐波分析的精度和分辨率,提出基于求根多重信号分类法（root-MUSIC)和自适应线性神经网络的间谐波参数估计方法.该算法利用求根多重信号分类法估计信号中谐波和间谐波的个数及频率,将谐波和间谐波的频率作为Adaline神经网络的输入进行学习,用得到的权值确定谐波和间谐波的幅值和相位；将频率作为权值在改进的Adaline神经网络中参与学习,估计谐波和间谐波的频率、幅值和相位.Matlab仿真结果表明,该算法频率分辨率高、检测准确、收敛快；当频率估计准确时,基本Adaline神经网络与改进的Adaline神经网络具有相近的检测精度,且前者的实时性更好.

Interharmonic parameter estimation based on Root-MUSIC and Adaline neural network

 An interharmonic parameter estimation algorithm based on root-multiple signal classification(MUSIC) and Adaline neural network was proposed in order to improve the analyzed precision and frequency resolution of the interharmonic in power system.  Root MUSIC algorithm was used to obtain the number and frequencies of the harmonics and interharmonics, and the frequencies were inputted into Adaline neural network for learning. Then the amplitudes and phases of the harmonics and interharmonics were estimated by using the weights. The frequencies were treated as the weights to be adjusted in the improved Adaline neural network in order to estimate the frequencies as well as the amplitudes and phases of the harmonics and interharmonics. The Matlab simulation results verified the  algorithm has the characteristics of high frequency resolution, detect accuracy and fast convergence. When the estimation of frequency is accurate, the Adaline neural network and its improved method have closely precision, and the real-time property of the former is better．