自适应线性神经网络LMM算法的谐波辨识技术研究

Research on harmonic identification technology of adaptive linear neural network LMM algorithm

介绍了组合适应线性神经网络最小平均值评估法(Adaline-LMM)对脉冲控制信号的拟合分析方法,用于对电力控制系统中的信号评估。通过对系统信号中的各个谐波分量的幅值和相位进行谐波辨识,并对Adaline的权重向量进行更新,同时对目标函数进行技术估计。其中,自适应神经网络中的权重向量由LMM算法进行迭代更新,通过最小平均值估计算法的引入,减小由于脉冲噪声引起的暂时波动的影响。通过对给定脉冲信号进行拟合,可以发现所提方法具有较高的计算精度。

The combined adaptive linear neural network minimum mean evaluation method(Adaline-LMM)is introduced in this paper,which can be used to perform the fitting analysis of the pulsed control signal and evaluate the signals in the power control system.The harmonic identification is conducted and the weight vector of Adaline is updated by means of the amplitude and phase of each harmonic component in the system signal.And also the target function is estimated.The LMM algorithm is used to conduct the iterative update of the weight vector in the adaptive neural network.The impact of temporary fluctuations caused by impulse noise is reduced due to the introduction of the minimum mean estimation algorithm.It is found that the method has high calculation accuracy by fitting a given pulse signal.