Electrical Power and Machines Department, Higher Institute of Engineering, El Shorouk Academy, Cairo, Egypt.
Electrical Power and Machines Department, Faculty of Engineering, Al Azhar University, Cairo, Egypt.
PLoS One. 2024 Oct 11;19(10):e0309926. doi: 10.1371/journal.pone.0309926. eCollection 2024.
Power transformers are essential elements in power systems and thus their protection schemes have critical importance. In this paper, a scheme is proposed for accurate discrimination and location of internal faults in power transformers using conventional measuring devices attached to the transformer. Different types of internal winding faults are intensely considered: partial discharge, inter-disk faults, series and shunt short circuit faults and axial displacement. Depending on the transformer measured output voltage, input voltage and the input current, the construction of a locus diagram (ΔV-Iin) serves as an indicator for any physical modification to the winding. Using five suggested features extracted from the developed locus, an artificial neural network (ANN) technique is applied to accurately distinguish any deviation from the transformer healthy condition. The exact location of each fault inside the windings of power transformer is then determined. The obtained results validate the usefulness of the proposed scheme for different internal faults. The superiority of the proposed scheme is extensively examined by comparing its results with some published schemes.
电力变压器是电力系统中的重要组成部分，因此其保护方案具有至关重要的意义。本文提出了一种利用附加在变压器上的常规测量设备准确区分和定位电力变压器内部故障的方案。本文深入考虑了不同类型的内部绕组故障，包括局部放电、相间故障、串联和并联短路故障以及轴向位移。根据变压器测量的输出电压、输入电压和输入电流，构造轨迹图（ΔV-Iin）作为绕组任何物理变化的指示。利用从所开发的轨迹中提取的五个建议特征，应用人工神经网络（ANN）技术来准确区分任何偏离变压器健康状态的情况。然后确定电力变压器绕组内部每个故障的确切位置。所获得的结果验证了所提出方案对于不同内部故障的有效性。通过与一些已发表的方案进行比较，广泛检验了所提出方案的优越性。
