基于小波和神经网络的矿用变压器保护装置的研究  

Research of Mine Used Transformer Microcomputer Protection Device Based on Wavelet and Neural Network

作　　者：公茂法[1] 殷凡姣 张超[1] 李美蓉[1] 顾强[1] 李杰[1] 

机构地区：[1]山东科技大学,山东青岛266590

出　　处：《煤矿机械》2015年第8期87-90,共4页Coal Mine Machinery

基　　金：山东省自然科学基金(ZR2012EEM021)

摘　　要：励磁涌流识别是矿用变压器微机保护中有待进一步解决的难题,鉴于传统判别原理的局限性,提出了一种基于小波和神经网络的励磁涌流识别方法,并结合比率差动保护原理,设计了以FPGA、DSP和ARM为核心的矿用变压器微机保护装置。小波和神经网络采用FPGA硬件实现方式,提高了保护的速动性。测试结果表明,该装置不误动且不拒动,能够正确区分励磁涌流和内部故障电流。Recognition of inrush current is still the key issue for mine used transformer protection. A new method based on wavelet and neural network is given considering the limitations of traditional discrimination principle. According to the method and ratio differential protection, a transformer microcomputer protection device is designed with FPGA, DSP and ARM. Wavelet and neural network are implemented on FPGA to improve the proteetion＇s speed. The test results prove that it can act reliably and identify the inrush current exactly.

关 键 词：小波变换 神经网络 励磁涌流 变压器 微机保护装置 

分 类 号：TM774[电气工程—电力系统及自动化]