基于改进Hausdorff距离算法的发变组大差保护新原理

发变组大差动保护属于保护双重化中主保护的一种,传统发变组大差保护动作速度慢且易受电流互感器饱和影响。对此,提出了一种兼备速动性及抗饱和特性的基于改进Hausdorff距离算法的发变组大差保护新判据。首先分析了改进Hausdorff距离算法的基本原理:在发变组区内故障时,由各端电流构造的Hausdorff距离值较大;在区外故障及正常运行时,由各端电流构造的Hausdorff距离值较小,基本为0。基于此原理提出了发变组大差保护新判据,并给出了门槛值整定方案。另外,提出了电流互感器饱和识别及保护再开放策略。理论分析和仿真结果验证了所提保护方案的有效性,能快速可靠识别区内故障,在区外故障饱和情况下可靠闭锁,再次发生转换性故障时也能快速开放,且判据门槛值整定简单,不受系统参数的影响。

New principle of generator-transformer large differential protection based on improved Hausdorff distance

The large differential protection of generator-transformer unit is one of the dual main protections. The traditional large differential protection of generator-transformer unit is easy to be influenced by current transformer saturation and the operation is slow. To dealing with above problems, a new criterion for large difference protection of generator-transformer unit based on improved Hausdorff Distance algorithm is proposed, which has high operation speed and anti-saturation capability. Firstly, the basic principle of the Hausdorff Distance algorithm is analyzed:when the fault occurs within the generator-transformer unit, the Hausdorff Distance constructed by the current of each terminal is big; when the fault is outside the generator-transformer unit or the system is under normal operation, the Hausdorff Distance constructed by the current of each terminal is smaller, which is basically 0. Based on this principle, a new criterion for the large difference protection of generator-transformer unit is proposed, and the threshold setting scheme is proposed. In addition, the current transformer saturation recognition strategy and protection opening strategy are proposed. Theoretical analysis and simulation results verify the effectiveness of the proposed protection scheme, which can quickly identify the internal fault and be reliably locked under external fault with current transformer saturation, can quickly open again under converted fault. The threshold value setting process is simple and not affected by system parameters. This work is supported by Science & Technology Program of Guangdong Power Grid Co., Ltd. (No. GDKJQQ20161176) and National Natural Science Foundation of China (No. 51677108).