电缆故障检测技术智能化的研究

发布时间：2018-03-18 23:27

本文选题：电缆故障　切入点：智能化　出处：《西安电子科技大学》2014年硕士论文　论文类型：学位论文

【摘要】：电力电缆已经成为现代工业文明的重要基础,逐渐取代了传统的架空线路,与工农业生产及我们的日常生活息息相关。然而无法避免的电缆故障带来的巨大经济损失已成为当今社会的一个严重问题,因此如何能在复杂的铺设环境中快速、有效、准确地检测出电缆故障点具有非常重要的现实意义。论文首先查阅大量文献资料介绍了电缆发生故障的原因以及故障类型,接着在分析对比了常用故障检测方法的基础上以时域脉冲原理作为本文的基本设计依据,然后简要介绍了电缆在长线模型下的行波传播理论以及基于该理论的行波传播速度、波阻抗以及行波的反射折射现象。在综合考虑了这些因素的前提下提出了电缆故障检测智能化中具有举足轻重作用的故障测试仪的具体技术指标以及相应的硬件实现电路和软件算法流程。测试仪的硬件设计主要以嵌入式单片机和CPLD构成的双控制器为核心,结合了脉冲发射模块、信号接收模块、高速ADC采样模块等单元模块;测试仪的软件算法包括单片机的嵌入式C语言设计以及CPLD的Verilog HDL硬件语言,C语言采用模块化设计思想而Verilog HDL采用自顶向下的基本设计结构。由于复杂的测试环境以及多种多样的干扰源,获得的实测波形往往包含很多波动不规则的震动,严重影响有效故障信息的识别以及后续的故障距离计算。正是基于这种背景本文又就回波处理智能化进行了初步研究,以小波分析为理论基础,通过对基于极大值原理、基于相关性原理以及阈值法三种经典的小波去噪法分析对比后,就阈值函数选择、阈值量化准则选择、小波基函数选择、小波分解尺度等影响阈值去噪效果的关键因素作了讨论,综合考虑之后选择db4作为小波基函数以及采用软阈值去噪法对故障波形进行智能化去噪处理。最后本文利用小波分析工具MATLAB对db4小波的实际去噪效果进行了仿真,并对在昆山市利用本文设计的智能故障测试仪测得的实地数据进行了去噪仿真处理,结果显示去噪效果明显。另外本文设计的测试仪也实现了检测电缆故障的功能并且测量误差在可接受范围内,具有很好的实用性。这表明本文对电缆故障检测智能化的研究取得了一定的成效。
[Abstract]:Power cables have become an important foundation of modern industrial civilization, gradually replacing the traditional overhead lines, It is closely related to industrial and agricultural production and our daily life. However, the enormous economic losses caused by unavoidable cable failures have become a serious problem in today's society, so how can they be rapid and effective in a complex laying environment? It is very important to detect the cable fault point accurately. Firstly, the paper introduces the reason and type of cable fault by consulting a lot of literature. Then, based on the analysis and comparison of common fault detection methods, the time-domain pulse principle is taken as the basic design basis of this paper, and then the traveling wave propagation theory and the traveling wave propagation velocity based on this theory are briefly introduced. The wave impedance and the reflection and refraction of traveling wave are taken into account. On the premise of synthetically considering these factors, the specific technical specifications and corresponding hardware of the fault tester which plays an important role in intelligent cable fault detection are put forward. The hardware design of the tester is mainly based on the dual controller composed of embedded microcontroller and CPLD. Pulse transmitting module, signal receiving module, high speed ADC sampling module and other unit modules are combined. The software algorithms of the tester include the embedded C language design of single chip microcomputer, the modular design idea of CPLD Verilog HDL hardware language C language and the top-down basic design structure of Verilog HDL. Because of the complex test environment, And a variety of sources of interference, The measured waveforms often contain a lot of irregular vibrations, which seriously affect the identification of effective fault information and the calculation of subsequent fault distance. Based on this background, this paper makes a preliminary study on intelligent echo processing. On the basis of wavelet analysis, after analyzing and comparing three classical wavelet denoising methods based on maximum principle, correlation principle and threshold method, the selection of threshold function, threshold quantization criterion and wavelet basis function are discussed. The key factors such as wavelet decomposition scale which affect the effect of threshold denoising are discussed. After comprehensive consideration, db4 is selected as wavelet basis function and soft threshold denoising method is adopted to deal with the fault waveform intelligently. Finally, the actual denoising effect of db4 wavelet is simulated by MATLAB, a wavelet analysis tool. The field data measured by the intelligent fault tester designed in this paper are simulated and processed in Kunshan. The results show that the denoising effect is obvious. In addition, the test instrument designed in this paper also realizes the function of detecting cable faults and the measurement error is within acceptable range. It has good practicability, which indicates that the intelligent research of cable fault detection in this paper has achieved some results.
【学位授予单位】：西安电子科技大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TM75

【相似文献】