基于动态云—量子神经网络群的配电网实时故障定位方法  

Real-time fault location method of distribution network based on dynamic cloud and quantum neural network group

作　　者：马亮[1] 杨萍萍[1] 高建宇[1] 

机构地区：[1]河北联合大学轻工学院,河北唐山063000

出　　处：《工矿自动化》2014年第11期71-75,共5页Journal Of Mine Automation

基　　金：河北省自然科学基金钢铁联合基金资助项目(F2012209015);河北联合大学轻工学院科学研究基金项目(qy20120012)

摘　　要：针对传统的配电网故障定位方法在配电网故障信号微弱时存在的故障数据交叉现象严重、实时性较差等问题,提出了一种基于动态云-量子神经网络群的配电网实时故障定位方法;构建了用于配电网故障定位的动态云-量子神经网络群结构模型,提出一种动态云-量子神经网络群改进算法,并给出了基于该算法的配电网实时故障定位步骤;在Matlab软件中采用该方法对某10kV配电网进行故障定位仿真研究,结果表明该方法能够实时、有效地实现故障信号微弱情况下的配电网故障定位,测试精度为97.39%,训练时间为0.001 6s。For resolving problems of serious fault data crossover phenomenon and poor real-time performance of traditional fault location methods of distribution network under the condition of weak fault signal,a real-time fault location method of distribution network based on dynamic cloud and quantum neural network group was proposed.A structure model of dynamic cloud and quantum neural network group was established for fault location of distribution network.An improved algorithm of dynamic cloud and quantum neural network group was proposed and real-time fault location steps based on the improved algorithm for distribution network were given.The method was simulated for fault location of a 10 kV distribution network with test accuracy of 97.39%and training time of 0.001 6s.The results show that the method realizes fault location of distribution network under the condition of weak fault signal realtimely and effectively.

关 键 词：配电网 故障定位 微弱故障信号 动态云QNN群 云理论 量子神经网络 

分 类 号：TD60[矿业工程—矿山机电]