Neural Network Theory In The Harmonic Measurement And Research

With the wide use of power electronic device, harmonic becomes serious gradually. Harmonic measurement is the basis of harmonic research. This article analyses harmonic measurement, stressing on the method of harmonic measurement. Based on the demonstration of the two popular methods-FFT and Instantaneous Reactive Power theory. Some improved methods are also proposed. Meanwhile th?application prospect of neural network in harmonic measurement is introduced. Some important problem about the application of neural network in harmonic measurement is discussed in detail. The ANN self-adaptive and muti-layer feedforward neural network applied to harmonic measurement are presented, and two corresponding harmonic circuit arc built to enhance efficiency of real time and accuracy. Simulation verifies the effectiveness of the methods proposed in this paper.