磁控开关型故障限流器神经网络自适应偏置电流控制  

Neural Network Based Adaptive Bias Current Control in Magnetic-Controlled Switcher Type Fault Current Limiter

作　　者：潘艳霞[1] 薛花[2] 

机构地区：[1]山西省电力公司电力交易中心,山西太原030001 [2]上海电力学院电力工程系,上海200090

出　　处：《供用电》2012年第1期5-9,共5页Distribution & Utilization

基　　金：国家自然科学基金项目(51007053)

摘　　要：针对磁控开关型故障限流器在故障后其偏置线圈中会产生偶次反电动势,三相全控整流桥的负载突变为带偶次反电动势的电阻电感负载,提出了基于神经网络的磁控开关型故障限流器自适应偏置电流控制系统。该系统采用两个神经网络来分别构成自适应控制网络和辨识网络,对控制器参数进行自适应调整和在线辨识,实现了偏置电流的实时控制。试验证明,利用该控制系统对限流器偏置电流进行控制,限流器的响应速度和限流效果可以得到保证。Due to the even frequency reverse electromotive force generated in the type fault current limiter （FCL） dc bias coil under the fault situation, the load of the three-phase full-bridge controlled rectifier becomes an impedance plus even frequency reverse electromotive force. This paper presents an approach of adaptive bias current control for magnetic-controlled switcher type FCL based on radial basis function（RBF） neural net- work. This method builds up an adaptive control neural network and an identification RBF neural network. The former is used to adaptively adjust the control parameters at running and the latter is used to offer on-line identification the controlled plant. The system can achieve the bias current real time control. The experimental results prove that the speed and current limiting effect can be guaranteed when using the proposed approach to control the FCL bias current.

关 键 词：故障限流器 神经网络 自适应控制 控制系统 

分 类 号：TM26[一般工业技术—材料科学与工程] TM474[电气工程—电工理论与新技术]