Koley Ebha, Verma Khushaboo, Ghosh Subhojit
Department of Electrical Engineering, National Institute of Technology, G.E. Road, Raipur, 492010 India.
Springerplus. 2015 Sep 25;4:551. doi: 10.1186/s40064-015-1342-7. eCollection 2015.
Restrictions on right of way and increasing power demand has boosted development of six phase transmission. It offers a viable alternative for transmitting more power, without major modification in existing structure of three phase double circuit transmission system. Inspite of the advantages, low acceptance of six phase system is attributed to the unavailability of a proper protection scheme. The complexity arising from large number of possible faults in six phase lines makes the protection quite challenging. The proposed work presents a hybrid wavelet transform and modular artificial neural network based fault detector, classifier and locator for six phase lines using single end data only. The standard deviation of the approximate coefficients of voltage and current signals obtained using discrete wavelet transform are applied as input to the modular artificial neural network for fault classification and location. The proposed scheme has been tested for all 120 types of shunt faults with variation in location, fault resistance, fault inception angles. The variation in power system parameters viz. short circuit capacity of the source and its X/R ratio, voltage, frequency and CT saturation has also been investigated. The result confirms the effectiveness and reliability of the proposed protection scheme which makes it ideal for real time implementation.
对线路通行权的限制以及不断增长的电力需求推动了六相输电的发展。它为传输更多电力提供了一种可行的替代方案，而无需对三相双回路输电系统的现有结构进行重大修改。尽管有这些优点，但六相系统的接受度较低归因于缺乏合适的保护方案。六相线路中大量可能故障所带来的复杂性使得保护工作颇具挑战性。所提出的工作提出了一种基于混合小波变换和模块化人工神经网络的故障检测器、分类器和定位器，用于仅使用单端数据的六相线路。使用离散小波变换获得的电压和电流信号近似系数的标准差被用作模块化人工神经网络进行故障分类和定位的输入。所提出的方案已针对所有120种类型的并联故障进行了测试，包括故障位置、故障电阻、故障起始角度的变化。还研究了电力系统参数的变化，即电源的短路容量及其X/R比、电压、频率和CT饱和情况。结果证实了所提出保护方案的有效性和可靠性，使其成为实时实施的理想选择。
