变压器噪声数字化分析及快速预警终端的研发

摘要

This paper introduces specific design for a kind of terminal for rapid prewarning on potential failure of transform-ers based on noise digital analysis.It firstly analyzes various physical quantities and selects sound sensor of faster physical re-sponse characteristic than that of temperature or gas sensor and sampling chip with Delta-Sigma technology for designing hardware.Secondly,it applies digital information processing technology for designing measuring analysis algorithm among which it uses infinite impulse response digital filter algorithm for calculating continuous equivalent sound pressure levels of Z, C&A frequency weighting and fast Fourier transformation algorithm for analyzing noise frequency spectrum.By simulating on noise information model and using experimental equipments for commissioning,it proves that this terminal is provided with higher measuring precision for realizing rapid prewarning on abnormal working conditions of transformers and suitable for real-time monitoring and protection on transformer noise of the smart grid.%介绍了一种基于噪声数字化分析能快速预警变压器潜伏性故障的终端的具体设计。首先,分析了变压器的各种物理量,选择比温度和气体传感器具有更快物理响应特性的声音传感器,以及Delta-Sigma技术的采样芯片来设计硬件；其次,应用数字信号处理技术设计测量分析算法,其中,使用无限脉冲响应(infinite impulse response,IIR)数字滤波器算法计算 Z、C、A频率计权的连续等效声压级,使用实序列快速傅氏变换(fast Fou-rier transformation,FFT)算法分析噪声频谱。对噪声信号模型进行仿真,并使用实验设备进行调试,仿真与实验结果证明该终端具有较高的测量精度,能达到快速预警变压器异常工作状态,适合于智能电网变压器噪声实时监测与保护的场合。