Abstract：Most of the grounding faults were caused by insulation deterioration in the distribution network. The process and classification of single-phase ground fault data benefits to improve the diagnosis of distribution network line’s fault status, thus presented the distribution network line insulation warning. In this paper, used the phase plane trajectory method to analyses the data of temporary grounding fault, through the division of the phase plane trajectory and extract the percentage feature of waveform in each region. Combining with the other features of fault, and take advantage of the neural network in the prediction, to forecast the number of days which insulation of distance line equipment is completely damaged. Through the example, demonstrates the feasibility of phase plane and genetic BP neural network. This method can effectively judge the deterioration degree of insulation and propose warning in time, and then ensure power supply continuity and efficiency.
林仁杰,李天友,蔡金锭. 基于相平面和遗传BP神经网络的配电网线路绝缘预警[J]. 电气技术, 2015, 16(9): 47-51.
Lin Renjie,Li Tianyou,Cai Jinding. The Warning of Insulation Deterioration based on the State Phase and Genetic BP Neural Network in Distribution Network. Electrical Engineering, 2015, 16(9): 47-51.
2015, Vol. 16 
