局部放电脉冲波形数字测量与分析系统

为检测局部放电快速变化的脉冲电流波形及根据波形进行放电模式识别工作 ,研制了一套数字测量与分析系统。系统硬件部分包括宽带信号检测器件、高速数字示波器等。实测着重考虑了高压测量中的屏蔽 ,高速采集系统的带宽 ,以及放电实验样品的多样性等因素。结果证明该测量系统具有灵敏度高 ,测量准确 ,对脉冲波形畸变小等优点 ,可用于测量局部放电脉冲波形。系统软件的采样程序能够通过人机界面完成对示波器的通信与控制 ,数据的人工或自动采集 ,波形绘制与打印 ,数据的二进制、ASCII码、图形文件保存及文件间转化。系统软件的数据处理程序能够完成对波形的检出和特征量的提取。特征提取中应用了包括自回归模型在内的多种方法。采用前馈神经网络对不同的放电模式进行了识别。研究结果表明 ,利用局部放电的脉冲波形可以识别不同模式的放电

Digital system on partial discharge pulse shape detection and analysis

A digital detection and analysis system was developed to record the partial discharge pulse and to recognize different patterns. The system hardware includes a wide band detector, a high speed oscilloscope, etc. Special consideration was given to the electromagnetic shielding, the system bandwidth, and a variety of discharge patterns. Testing results have shown its sensitivity, accuracy, and small distortion for all waveforms, all of which contribute to its ability to extract transient pulse signals. The software includes data acquisition and data processing programs. The former is designed to accomplish communication and oscilloscope controll, automatic sampling, waveform drawing and printing. The data can be saved as binary, text, or graphic files. The data processing program can check the waveform, and extract features. Auto regression method is employed for feature extraction. Artificial neural network is used to recognize different patterns. The result shows the potential of applying waveform detection in partial discharge pattern recognition.