基于改进神经网络的10 kV配电网相间短路故障自动化检测系统

Automated Detection System for Interphase Short Circuit Fault in 10 kV Distribution Network Based on Improved Neural Network

为提高10 kV配电网相间短路故障自动化检测精度,设计了一种基于改进神经网络的10 kV配电网相间短路故障自动化检测系统。该系统由在线监测模块的探头装置,检测10 kV配电网线路电压、电流信号信息,通过短距离无线网与通信模块GPRS网络连接,发送至数据分析模块;数据分析模块由服务器、计算机端构成,电压、电流信号数据会实时缓存于服务器中,由计算机端使用基于改进神经网络的相间短路故障诊断模型,诊断电压、电流信号样本,是否属于10 kV配电网相间短路故障状态,完成10 kV配电网相间短路故障自动化检测。通过实验验证,该系统在10 kV配电网相间短路故障检测方面具有很高的准确性,无须人工操作,具有广泛的应用前景。

To improve the accuracy of automatic detection of phase to phase short circuit faults in 10 kV distribution networks,an improved neural network based automatic detection system for phase to phase short circuit faults in 10 kV distribution networks is designed.The system consists of a probe device in the online monitoring module,which detects the voltage and current signal information of the 10 kV distribution network line.It is connected to the communication module GPRS network through a short distance wireless network and sent to the data analysis module.The data analysis module is composed of a server and a computer.The voltage and current signal data are cached in real-time in the server.The computer uses an improved neural network-based interphase short circuit fault diagnosis model to diagnose voltage and current signal samples,determine whether they belong to the 10 kV distribution network interphase short circuit fault state,and complete the automatic detection of 10 kV distribution network interphase short circuit faults.Through experimental verification.This system has high accuracy in detecting phase to phase short circuit faults in 10 kV distribution networks,without manual operation,and has broad application prospects.