摘要
当前的电力通信网在故障诊断方面只能进行定性的故障诊断与分析,无法实现复杂通信环境下定量的故障诊断与分析,因此将神经网络方法应用于电力通信网的故障诊断中,以实现定量的故障诊断与分析.首先,确定电力通信网故障信息量化计算指标,构建电力通信网的故障危害性评价体系,并引入专家评估系统;通过专家评估系统处理电力通信网上的故障信息;引入神经网络方法,对电力通信网中的权重参数进行训练,构建基于神经网络的电力通信网故障定量分析模型,实现对故障的量化分析与处理.通过算例测试表明,所提方法能有效计算出电力通信网故障等级信息及定量报警值,与传统方法相比更先进,可进行大规模推广应用.
Abstract
China's current electric communication network can only carry out qualitative rather than quantitative fault diag-nosis and analysis under complex communication environment. Therefore this paper studies the application of neural net-work method to fault diagnosis of electric communication network,attempting to realize quantitative fault diagnosis and a-nalysis.First the quantitative calculation index of fault information of electric communication network is determined,the fault hazard evaluation system is constructed,and the expert evaluation system is introduced.Then the fault information is processed by the expert evaluation system,and neural network method is adopted in training of weight parameters,ena-bling modeling of neural network-based quantitative fault analysis and realizing quantitative fault analysis.The proposed method is verified by case test effective in calculating fault level information and quantitative alarm value of electric com-munication network,and thus more advanced than conventional methods and worthy of large scale application.
NETL
NSTL
维普
万方数据