王新胜.Ip-iq谐波检测中三角基函数神经网络快速滤波方法的研究[J].电测与仪表,2014,51(18):. wang xin-sheng.Study of the fast filtering based on trigonometrically-activated neural networks in the ip-iq harmonic detection[J].Electrical Measurement & Instrumentation,2014,51(18):.
Ip-iq谐波检测中三角基函数神经网络快速滤波方法的研究
Study of the fast filtering based on trigonometrically-activated neural networks in the ip-iq harmonic detection
Most of ip-iq harmonic detection of the instantaneous reactive power based on the low-pass filter (LPF), in order to achieve stable output, it need to collect large amounts of data, the response is associated with the amount of data. Many of the existing harmonic analysis using neural network, but it is based on the data directly into already exists, the real-time cannot reflect. In this paper advanced the real-time detection method of the trigonometrically activated neural networks instead of LPF based on the above two methods’s defect , Not in the analysis of numerical filter fixed ,write a S function in Simulink and encapsulated into modules, shorten the response time, reduce the relative error; By comparing the MATLAB simulation results, it show that the trigonometrically-activated neural networks can completely replace the low pass filter, quick response, the relative error of relatively small, and stable output.