基于混合遗传算法的变压器局部放电超声定位法  

Ultrasonic localization method of partial discharge in power transformer based on hybrid genetic algorithm

作　　者：刘化龙[1] 胡钋[1] 青志明[2] 

机构地区：[1]武汉大学电气工程学院,湖北武汉430072 [2]国网重庆市电力公司技能培训中心,重庆400053

出　　处：《兰州理工大学学报》2014年第6期105-109,共5页Journal of Lanzhou University of Technology

基　　金：国家自然科学基金(51177112)

摘　　要：论述变压器局部放电超声波定位法原理,把混合遗传算法用于局部放电的超声定位,从而提出基于混合遗传算法的变压器局部放电超声定位法;并将其定位结果与遗传算法、模拟退火算法、粒子群算法和线性粒子群算法等智能算法的定位结果进行比较.算例结果表明混合遗传算法能有效地防止结果陷入局部最优,其定位效果理想,且比上述智能算法的定位效果都好.The principle of ultrasonic localization method of partial discharge in power transformer is ex- pounded and the hybrid genetic algorithm is applied to the ultrasonic localization, so that an ultrasonic lo- calization method of partial discharge in power transformer is presented based on hybrid genetic algorithm. The result of its localization is compared with that of other localization methods such as genetic algorithm, simulated annealing algorithm, particle swarm optimization （PSO） algorithm, and linear particle swarm optimization algorithm. It is shown by the results of calculation example that the hybrid genetic algorithm can effectively prevent the localization from being caught in local optimal, making the effectiveness of lo- calization with hybrid genetic algorithm ideal and better than all other intelligent algorithms mentioned a- bove.

关 键 词：局部放电 电力变压器 超声波定位 混合遗传算法 智能算法 

分 类 号：TM855[电气工程—高电压与绝缘技术] TM930.12