基于粒子群优化神经网络的谐波检测

Harmonic detection of neural network based on particle swarm optimization

根据谐波的傅里叶分析,把对谐波相位和幅值的检测转化为对谐波的正余弦分量幅值的检测.提出一种应用基于粒子群优化算法的多层前馈神经网络(MLFANN)实现谐波检测的方法,并构造一个3层MLFNN,以电网中最常见的3次、5次谐波为例,给出检测的实现方法.MATLAB仿真结果表明,该谐波检测方法具有较强的泛化能力和较高的检测精度.

Amplitude and phase are key parameters of harmonic by the Fourier analysis,the measurement of amplitude and phase of harmonic can be translated into measurement of amplitude of its sine and cosine component.In this paper we put forward a method to measure harmonic by using a multi-layer feed forward neural network(MLFNN)which is based on the particle swarm optimization,and construct a MLFNN of three layers,using the 3rd and 5th harmonic as an example to expatiate the realization of this measure-method in practice.Using Matlab tools,we proved the method has strong capacity of generalization and relatively high precision.