改进人工蜂群优化Ostu电力设备红外图像分割算法  

Power equipment infrared image segmentation algorithm based on improved artificial bee colony optimization Ostu

作　　者：冯新宇 付志伟 柴侨峥 李亚妮 刘晓磊 Feng Xinyu;Fu Zhiwei;Chai Qiaozheng;Li Yani;Liu Xiaolei(School of Electrical&Control Engineering,Heilongjiang University of Science&Technology,Harbin 150022,China;State Grid Heilongjiang Electric Power Co.Ltd.,Daxing'anling Power Supply Company,Daxing'anling 165000,China)

机构地区：[1]黑龙江科技大学电气与控制工程学院,哈尔滨150022 [2]国网黑龙江省电力有限公司大兴安岭供电公司,黑龙江大兴安岭165000

出　　处：《黑龙江科技大学学报》2023年第6期909-916,共8页Journal of Heilongjiang University of Science And Technology

基　　金：国家电网公司2023年科技项目(SGHLXA00SZJS2310073);黑龙江省省属本科高校基本科研业务费项目(2021-KYYWF-1477)。

摘　　要：针对红外电力设备检测系统中红外图像分割效果差、易陷入局部最优等问题,提出一种改进人工蜂群优化Ostu电力设备红外图像分割算法。在种群初始化阶段融合了Lévy飞行增强蜂群的空间搜索能力和种群多样性,提高算法寻优的过程;在引领蜂阶段采用天牛须搜索法改进蜂群的搜索方程,提高蜂群的全局搜索能力,避免陷入局部最优。结果表明,与Ostu、PSO-Ostu和ABC-Ostu算法相比,文中算法的分割召回率和重叠率均提高了10%以上,有效提升了电力设备红外图像的分割效果。This paper is intended to address the poor effect of infrared image segmentation and falling into local optimum in infrared power equipment detection system and proposes an improved Ostu power equipment infrared image segmentation algorithm with artificial swarm optimization.The study involves improving the process of algorithm seeking by integrating Lévy flight in the initialization stage of the population to enhance the spatial search ability and population diversity of the swarm;and improving the global search ability of the swarm and avoiding falling into local optimum by using the aspen whisker search method to improve the search equation of the swarm at the leading stage.The experimental results show that compared with Ostu,PSO-Ostu and ABC-Ostu algorithms,the segmentation recall and overlap rate with the algorithm in this paper are improved over 10%,which effectively improves the segmentation effect of infrared images of power equipment.

关 键 词：电力设备 图像分割 人工蜂群 

分 类 号：TM216[一般工业技术—材料科学与工程] TP391.41[电气工程—电工理论与新技术]