基于混沌控制和多重滤波的局部放电去噪研究

De-noising of partial discharge based on chaotic control and multiple-filter

采集局部放电信号时,经常受到外界的干扰,主要包括周期性窄带干扰、白噪声干扰和随机脉冲干扰。噪声太大时,局部放电信号会被淹没。为了提高变压器局部放电信号定位的准确性,提出混沌控制和多重滤波的方法,在线检测去除干扰。利用相关性和谱分析识别周期干扰并确定窄带干扰的周期,设置合适参数控制混沌系统,精确去除噪声,采用多重滤波的方法去除白噪声干扰,以及利用脉冲极性鉴别法抑制随机脉冲干扰,并在虚拟仪器平台上进行仿真实验。实验结果表明,该方法可以有效抑制干扰,为变压器局部放电信号的定位系统奠定了技术基础。

The collection of partial discharge signals is often disturbed by the environment,which includes periodic narrow-band interference,white-noise interference,and random pulse shaped interference.When the noise is very large,partial discharge signals will be over whelmed.In order to improve the accuracy of location of partial discharge signals,this paper describes the method to no-line detect and remove the disturbance based on chaotic control and MFT.Using relativity and spectrum analysis to identify the periodic disturbance and determine the periodic of narrow bandwidth interference,setting appropriate parameter to control chotic system to denoise accurately.It proposes a multiple filter to remove white-noise and uses pulsic polarity distinguishing method to suppress random pulse noise and conduct a simulation based on a virtual instrumentation.The result shows that this method can suppress noise effectually.It lays a technique foundation for the location system of partial discharge signals of transformers.