基于神经网络的自适应谐波和无功电流的检测

Adaptive Detecting of Harmonic and Reactive Currents Based on Neural Network

提出了一种基于神经网络的自适应谐波和无功电流的检测方法,详细分析了这种方法的原理,给出了这种方法的具体实现如神经网络参考输入的选取、权值的更新等并讨论了参考输入的选取对测量结果的影响。仿真结果证明了该检测方法可以自适应跟踪负载电流的变化,准确检测谐波和无功电流,可用于有源电力滤波器中。

It proposes an adaptive method of detecting harmonic and reactive currents based on neural network,analyzes the principle of the method,gives its realization such as selections of the referenced inputs of the neural network and updating of the neural network's weight and so on and discusses influence on detecting results of selections of the referenced inputs.The results of simulation confirm that the detecting method can track the variation of the load current adaptively and detect harmonic and reactive currents accurately,which and be applied to active power filter.