摘要
标准遗传算法(SGA)经常早熟并且收敛慢,在用于粗糙集属性约简时,往往只能得到部分最小条件属性组合,并且很难收敛。针对传统遗传算法的这些缺陷,采用多种群遗传算法(MPGA),用移民算子、局部竞争法来保持种群的多样性,以防止其陷入局部最优。将多种群遗传算法用于粗糙集属性约简,可得到所有的最小条件属性组合,并且收敛速度快。实例应用验证了该算法的有效性,可用于变压器的故障诊断。
Standard genetic algorithm（ SGA） is always early- maturing and slowly converging; what＇s more,it gets only part of the minimum condition attribute combination and is difficult to converge when used for rough set attribute reduction. In order to solve these problems in traditional genetic algorithm,the multiple population genetic algorithm（ MPGA） is used. Immigration operator and local competition are used to maintain the diversity of population and avoid falling into local optimal value. When MPGA is used for rough set attribute reduction,it can fast get all the minimum combination of condition attributes and converge in high speed. The practical examples verify that the algorithm is effective,and can be applied to the fault diagnosis of transformer.
出处
《自动化仪表》
CAS
2016年第4期27-30,共4页
Process Automation Instrumentation
基金
上海市电站自动化技术重点实验室基金资助项目(编号:13DZ2273800)
关键词
人工智能算法
标准遗传算法
故障诊断
变压器
粗糙集
多种群
移民算子
神经网络
Intelligence artificial algorithm
Standard genetic algorithm（SGA）
Fault diagnosis
Transformer
Rough set
Multiple population
Immigration operator
Neural network
