基于小波神经网络单相断线故障选线和定位

Single-phase open-line fault selection and location based on wavelet neural networks

根据发生单相断线不接地故障、单相断线电源侧接地故障或单相断线负荷侧接地故障时,正序电流暂态分量的变化特性,提出了一种配电网单相断线故障选线和定位保护新方法。该方法应用小波奇异性检测功能,获取故障引起的正序电流暂态分量模极大值的极性和大小,进行故障选线;并和小波神经网络相结合实现模极大值与故障点位置之间的映射关系,进行故障定位。仿真结果表明,该故障选线和定位方法准确可靠。

According to the change characteristic of the transient component of positive-sequence current when single-phase open-line fault with non-grounding or power-side-grounding or load-side-grounding happens,a novel selection and location method for single-phase-open-line fault of distribution network is presented.In this method,singularity detection of wavelet is employed to obtain the polarities and magnitudes of modulus maximum of the transient component of positive-sequence current caused by the fault.Comparing them,the fault line can be determined.The polarities and magnitudes are combined with wavelet neural networks to realize the mapping relationship between modulus maximum and fault point position,and then the fault location can be realized.Simulation results show that the proposed method is of precision and reliability.