基于AR谱估计和Adaline神经元的间谐波分析  

Inter-harmonic Analysis Based on AR Spectral Estimation and Adaline Neural Network

作　　者：蔡忠法[1] 陈隆道[1] 

机构地区：[1]浙江大学电气工程学院,浙江省杭州市310027

出　　处：《电力系统自动化》2007年第17期78-82,共5页Automation of Electric Power Systems

摘　　要：将人工神经网络应用于电力系统间谐波分析,提出了Burg算法与Adaline神经元相结合的间谐波分析方法。通过Burg算法计算自回归(AR)模型参数,得到信号中谐波和间谐波的个数及频率初值,然后应用改进的Adaline神经网络精确分析谐波和间谐波的频率、幅值和相位。MATLAB仿真结果验证了Burg-Adaline间谐波分析方法具有分辨率高、精度高、收敛快的优点。An inter-harmonic analysis algorithm based on Burg spectral estimation and Adaline neural network is proposed to analyze the harmonic and inter-harmonics in power systems. In the Burg-Adaline inter-harmonic analysis approach, Burg algorithm is used for calculating the parameters of AR model to obtain the pre-estimated frequencies and the number of the harmonics and inter harmonics. Exact frequencies, amplitudes and phase angles are given by an improved Adaline neural network. The MATLAB simulation results verify the validity of the proposed algorithm with the characteristics of high resolution, superb precision and fast convergence.

关 键 词：电能质量 间谐波 AR模型 Burg算法 神经网络 ADALINE 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程] TM711[自动化与计算机技术—控制科学与工程]