摘要
本文提出结合运用BP神经网络和GRNN神经网络建立的故障诊断模型与模糊决策技术的电网故障诊断方法。首先,该方法对电网的三类主要元件(变压器、线路和母线)分别建立神经网络模型。其次,该方法中元件的神经网络基于电力系统中继电保护装置信息和断路器的状态信息进行初步诊断。神经网络模型中BP神经网络负责将该元件的状态和信息进行预处理和判断,将结果反馈给广义回归神经网络GRNN;GRNN神经网络负责在发生故障以及电网拓扑结构发生变化的情况下准确找到故障源。最后,专家系统根据得到的初步诊断信息运用模糊决策技术进行综合诊断。经本文分析及测试,该方法能够有效的提高运行人员故障处理效率,提高电力系统供电的可靠性和安全性。
This paper proposes a method of a combination of BP neural network and GRNN neural network fault diagno- sis model and fuzzy decision method of fault diagnosis technology. First, the three main components of the grid （trans- formers, lines and busbars） were established by the neural network model. Then component neural network have a pre- liminary diagnosis based on power system protection devices and breaker status information. A neural network model of BP neural network responsible for the state of the elements and in-formation preprocessing and judgment, the results back to the generalized regression neural network GRNN;GRNN neural network is responsible in the event of failure, and net- work topology changes, the situation un-der accurately find fault source. Finally expert system based on diagnostic infor- mation obtained in the initial deci-sion-making techniques using fuzzy comprehensive diagnosis. The paper analyzes and testing, this method can effectively improve the efficiency of operating personnel troubleshooting, improve power system reliabil-ity and security.
出处
《电气开关》
2018年第1期87-90,93,共5页
Electric Switchgear
关键词
神经网络
专家系统
电网
neural networks
expert system
power grid
