【作者】 林金洪；

【导师】 熊小伏；

【作者基本信息】 重庆大学 ， 电气工程， 2003， 硕士

【摘要】 随着电缆应用成本的下降和城市电网改造工作的开展，电缆获得了越来越广泛的应用，但是，到目前为止，电力电缆故障测距仍然缺少有效的方法。本文在现有的电缆故障测距方法的基础上，通过在原理和方法上进行深入研究，提出了三种在原理上更精确、工程上更实用的电缆故障测距新方法。基于小波重构的电缆故障测距方法采用脉冲电源作用下故障相与健全相的电流差作为测量信号，利用小波变换对其作多尺度分解，然后对信号在高频下进行单支重构。与传统的行波测距方法相比，这种方法不受电缆分支接头或其他阻抗不匹配点反射波的干扰，不受故障类型的影响，在近区也不存在无法识别反射波的问题，同时也减少了波速不确定性对测距精度的影响。第二种方法将小波变换与模式变换理论结合起来，采用的是暂态行波信号，首先将三相信号转换成模式分量，零模分量的小波变换系数用于判别故障的大致位置，然后利用线模分量的小波变换系数来确定行波到达时间。这种方法由于采用了模式变换，避免了传统行波方法中存在的受故障起始角的影响的问题。第三种方法采用电缆线路的分布参数线路模型，根据功率平衡原理推导出故障测距方程，并通过搜索迭代方法解之。这种方法仅采用线路的单端数据，不受故障过渡电阻的影响，无需求解复杂的长线方程。本文利用EMTP和MATLAB程序对上述方法进行了数值仿真，结果表明：(1) 在本文所提出的基于小波重构的电缆故障测距方法中，利用健全相和故障相的电流差信号进行故障定位，能消除故障点之外其他阻抗不匹配点的反射波的影响。通过对信号进行小波变换，可解决行波到达时间的确定和死区方面的问题。(2) 通过对电流进行小波变换并在高频下进行单支重构，可以减小行波波速不确定性对测距的影响。(3) 利用线模电流的小波变换系数进行测距，不仅不受故障电阻的影响，也不受故障起始角的影响。(4) 利用功率平衡理论进行故障测距能够在原理上克服传统阻抗法普遍存在的受故障电阻影响的缺陷。 (5) 本文所提出的方法有很高的测距精度，方法是可行的。更多还原

【Abstract】 With application costs decreasing of power cable and rebuilding of power network, power cable is being widely used. But, until now, there have not efficient methods on power cable fault location. Through research on principle and approach, this paper puts forward three new methods in power cable fault location. The first method is based on wavelet reconstruction, and the measured signal which is the difference between the faulty and sound phase current under the high voltage pulse excitation, is transformed using wavelet, and the high frequency component is reconstructed at single scale. Compared with conventional methods, the proposed method is not disturbed by reflected waves from tee joints or other resistance discontinuities, and is not influenced by fault type, and can recognize the reflected wave from the fault point even when the fault point is near to the measuring terminal (so-called dead-zone), and simultaneously, the method reduces the influence of wave speed uncertainty. In the second method, wavelet transform is combined with modal transform, and transient travelling-wave signal is used. After three phase signals are decomposed into their modal components, the wavelet transform coefficient of ground mode can be used to identify approximate position of fault, and the wavelet transform coefficient of aerial mode is used for identifying arrival time of traveling-wave. Because the method utilizes modal transform, the problem of fault inception angle can be avoided , which occurs in almost all conventional methods. The last proposed method is based on the distributed parameter model, the algorithm utilizes searching and iteration methods to locate the fault point. Only using data of single-terminal, the algorithm can eliminate the influences of the fault resistances and it needn’t resolve the complicated equation of long line. The methods above have been simulated using EMTP and MATLAB, and the simulation results indicate that, (1) In the fault location method based on wavelet reconstruction, the influences of reflected waves from tee joints or resistance discontinuities except the fault point can be eliminated when the current difference between faulty phase and sound phase is used in fault location. And the problem of wave arrival time and dead-zone can be resolved through wavelet transform of current signal.(2) The influences of wave propagation speed uncertainty can be reduced when the measured current is decomposed using wavelet transform, and the high frequency<WP=6>component is reconstructed at single scale.(3) The location result is not influenced by fault resistance and inception angle when wavelet transform coefficient of aerial mode current is used in fault location.(4) The location results are always affected by fault resistance in almost all conventional impedance methods in fault location, but the application of power balance theory can overcome the shortcoming.(5) The proposed methods have high accuracy in fault location, and the methods are feasible.更多还原