【作者】 陈双；

【导师】 何正友；

【作者基本信息】 西南交通大学 ， 电力系统及其自动化， 2013， 硕士

【摘要】 随着电网的加速建设,高压输电网的枢纽作用日渐凸显,快速准确地进行故障点定位以及时恢复供电,具有重要的社会效益和经济价值。本文使用的基于行波固有频率的故障定位方法较之现有基于时域行波的方法,最大优点是避免了波头识别困难的问题,并能准确计算行波波速。但该方法在输电网实际应用中还存在测量端系统等效阻抗难以准确获取,输电网中多端系统的故障支路定位等问题。钊对系统等效阻抗对故障定位的影响,本文提出了基于行波固有频率的故障定位中测量端系统等效阻抗识别方法。首先将测量端系统电阻与电感视为待识别参数,利用故障行波多次固有频率信息构造非线性多目标函数,再由Nelder-Mead单纯性搜索算法求解局部最优值,得到满足故障定位要求的等效阻抗。因为目标函数具有多解性,满足局部最优的等效阻抗不代表系统实际阻抗值,但由此能计算正确的故障行波系统端反射角,最终得到准确的故障距离,形成规避系统等效阻抗影响的行波固有频率故障定位改进方法。该方法无需增加其他电气量,且和原有的定位方法计算形式保持一致,具有良好的可移植性。大量仿真试验验证表明定位结果不受故障距离、故障类型和过渡电阻的影响,且迭代算法对迭代初值不敏感。另外,在故障距离计算中能实现对定位结果准确度的评价,显著增加原有方法的可靠性,避免提取错误的固有频率而导致定位失败。针对输电网通常出现的T型结构,本文提出了基于行波固有频率的故障定位在T型线路的解决方案。本文通过分析在T型线路不同位置发生故障时,母线测量电流所反映的行波传播路径,揭示出故障支路母线所测固有频率具有规律的谐波特性,并进一步数学推导出了故障行波在不同系统边界条件下,于故障点和对应母线间传播的固有频率变化趋势,并以此特性构造故障支路判别判据。在判断故障支路后则提取该支路母线电流,利用现有的单端固有频率定位方法计算故障距离,实现完整的T型线路故障定位。仿真实验验证表明定位结果不受故障距离、故障类型等因素的影响,特别在T节点附近发生单相接地故障时,亦能可靠判断故障支路。故障定位的死区出现在母线附近,这是由行波固有频率定位原理本身引起的。本文解决了行波固有频率定位方法对准确系统等效阻抗的依赖问题,同时以数学表达式揭示了故障行波的固有频率变化趋势,实现了可靠的T型电线路的故障支路判别和定位,以上改进均为该方法在输电网故障定位的实际应用奠定了一定的理论基础。更多还原

【Abstract】 With the accelerated construction of power grid, it is becoming increasingly prominent for high-voltage power transmission network being the pivotal role, which has important economic value and social benefits to quickly and accurately locate fault and restore power. In this paper, the fault location algorithm based on natural frequencies of traveling wave, compared with the existing traveling wave method based on the time domain, has most notable advantage in avoiding the difficult problem of wave-head identification, and can accurately calculate the wave velocity. But the practical application of this method in transmission network exists several problems, such as the difficulty to obtain accurate equivalent impedances in local system, and the problem of fault branch location in multiterminal system of transmission network.In view of the influence of system equivalent impedance to the fault location, this article proposed the impedance identification method of measuring end in fault location based on natural frequencies of traveling wave. Firstly, the local system resistance and inductance are considered as parameters to be identified, then multi-objective nonlinear functions are constructed using multiple natural frequencies of fault traveling wave, again by Nelder-Mead, namely simplex search algorithm, to solve the local optimal value to obtain the equivalent impedance satisfying the requirement of fault location. Because of the multiple solutions in multi-objective functions, the obtained local optimal equivalent impedance is not representative of the actual system impedance value, but it can correctly calculate the reflection angle of fault traveling wave, and finally get accurate fault distance, which make improvements to fault location method based on traveling wave natural frequencies in evading the influence of system equivalent impedance. This method eliminates the need to add other electrical parameters, and consistents on the form of original location algorithm, which has good portability. A large number of simulation experiments verify that the fault location results are not affected by the change of fault distance, fault type and transition resistance, and the iterative algorithm is not sensitive to the initial iterative. In addition, it can realize evaluation of location accuracy during the fault distance calculation, also can significantly increase the reliability of the fault location method based on the natural frequencies of traveling wave, to avoid extract the error natural frequencies.For T type structure usually appearing in transmission network, this paper presents the fault location based on traveling wave natural frequencies in the T type line solution. In this paper, through the analysis of the wave propagation path reflected by busline current considering different fault positions in T-line, it is revealed that the natural frequencies measured by the fault branch busline has a regular harmonic characteristic. Moreover, further study of the natural frequencies of the spread between the fault point and busline has been proofed by mathematical derivation of the trend of fault generated traveling waves in different system boundary conditions, with which the fault discrimination criterion has been structured. After determining the fault branch, extract the branch busline current, then use the existing single-end fault locating method based on natural frequencies to calculate fault distance, finally the complete T-line fault location method can be achieved. Simulation results show that the location results are not affected by the fault distance and the fault type, especially the fault branch can be reliably determined when single-phase faults near the T-node. Fault location dead zone is near the proximal busline, which is caused by the traveling wave frequencies positioning principle itself.This article addresses the traveling wave fault location method based on natural frequencies dependent on the accurate system equivalent impedance, meanwhile mathematical expressions are proposed to reveal the trend of natural frequencies of the traveling wave, and to achieve reliable fault branch discrimination and fault location in T-line. These above improvements are laid a theoretical foundation for the practical application of the method for fault location in transmission network.更多还原