基于两种混合智能算法对变压器铁心截面设计的优化研究  

Application of Two Hybrid Intelligent Algorithms to Research on Optimization of Design for Transformer Core Section

作　　者：李正绪[1] LI Zheng-xu(Siemens Transformer （Guangzhou） Co., Ltd., Guangzhou 510530,China)

机构地区：[1]广州西门子变压器有限公司,广州510530

出　　处：《电力学报》2016年第6期472-479,共8页Journal of Electric Power

摘　　要：为了降低电力变压器制造成本,对实际大型铁心的设计原则进行了研究,考虑了油道、分裂片对铁心截面的影响。结合人工智能算法,提出了采用粒子群、基因遗传并结合邻域搜索的两种混合算法对铁心截面进行优化设计。通过对铁心模型建立特定的邻域解及邻域范围来实现邻域搜索,并分别结合粒子群、改进的基因遗传算法流程,建立新型电力变压器铁心截面优化设计方案。具有良好的优化效果。特别是应用在大型电力变压器铁心设计制造中,具有更多的成本节省空间。In order to reduce the manufacturing cost of power transformer,according to the actual design principle of large core,considering the impact of the oil space and split sheets on core section.Combined with artificial intelligence algorithm,a hybrid algorithm based on particle swarm optimization and improved genetic algorithm combined with neighborhood search is proposed to optimize the core section.To realize the neighborhood search on core section model,and respectively combined with the algorithm flow of particle swarm optimization and improved genetic algorithm.Establishing a new type of optimization design,and achieves well optimization results.With more cost-saving space especially in the application of large transformer core design.

关 键 词：电力变压器 铁心截面 粒子群 基因遗传 邻域搜索 

分 类 号：TM41[电气工程—电器]