潜在等式约束的配电网遗传算法故障定位  

Genetic Algorithm Fault Location of Distribution Network with Latent Equality Constraint

作　　者：郭壮志[1] 陈波[1] 刘灿萍[1] 成蓬勃[1] 

机构地区：[1]广西大学电气工程学院,广西南宁530004

出　　处：《现代电力》2007年第3期24-28,共5页Modern Electric Power

摘　　要：如何建立遗传算法故障诊断数学模型,是提高配电网故障定位准确性和算法容错性能的基本前提。通过将一个由设备状态信息确定的多变量等式约束条件,经不同的等价变换,隐含于开关函数中,建立一个新的故障诊断数学模型。由于潜在等式约束条件的存在,避免了寻求最优解存在一值多解的可能性,能够实现配电网在有无信息畸变情况下单一故障的精确定位,具有很高的容错性。仿真结果证明了该算法的有效性。The construction of fault diagnosis mathematical model of genetic algorithm is the basic premise of improving the accuracy and fault-tolerant performance of distribution network fault location. A new fault diagnosis mathematical model is established. A multivariable equality constraint condition is hidden in switching functions through different equivalent transformations. The possibility of multiple solutions in the optimal solution is avoided for the existing of latent equality constraint condition. The accurate single fault location of distribution network is realized in the information distortion situation. This algorithm has high fault tolerance. Simulation results prove its validity.

关 键 词：遗传算法 等式约束 配电网 故障定位 

分 类 号：TM711[电气工程—电力系统及自动化]