摘要
针对电力设备红外图像批量诊断中故障特征参量提取及参数配置难题,采用粒子群算法(PSO)与Niblack算法相结合的方法,将设备热像从背景中分割出来并提取出设备的最低、最高及平均温度等参量,通过计算设备各温升特征,构建支持向量机(SVM)样本特征空间。采用优化的蝙蝠算法(BA)对SVM参数进行寻优,并利用最优参数配置下的SVM实现设备故障诊断。对220组图像样本测试结果表明:该红外图像故障诊断方法在电力设备热故障缺陷检测方面的效率及准确率较高,适用于电力大数据中非结构化红外图像的批量分析与处理。
Aiming at the problem of defect test and parameter assignment in the batcinfrared image ,PSO and Niblack algorithm are used to separa te the equipment therextract the lowest,highest and average temperature. Then,the SVM sample feature space can be constructed by calculat ing the tempe rature rise characteristics of the equipment The support vector machine (SVM) parameters are op ti-mized by using the optimized bat algorithm( BA),and the equipment defects diagnosis is realized by SVM under theoptimal parameter configuration. According to the 220 groups of image sample testing resuhigh efficiency and accura cy in thermal defects detection of power equipment,and is sui table processing of unstructured infrared images in large power data.
作者
李鑫
崔昊杨
许永鹏
李高芳
秦伦明
LI Xin;CUI Hao-yang;XU Yong-peng;LI Gao-fang;QIN Lun-ming(Department of Electronics and Information Engineering,Shanghai University of Electric Power,Shanghai 200090,China;Department of Electrical Engineering,Shanghai Jiao Tong University ,Shanghai 2002 40,China)
出处
《激光与红外》
CAS
CSCD
北大核心
2018年第5期659-664,共6页
Laser & Infrared
基金
国家自然科学基金资助项目(No.61107081
No.11647023)
上海市科委地方院校能力建设项目资助课题(No.15110500900
No.14110500900)
上海市自然科学基金面上项目(No.17ZR1411500)资助
关键词
故障诊断
粒子群算法
Niblack算法
支持向量机
蝙蝠算法
fault diagnosis
p a rtic le swarm optimization
Niblack algorithm
support vector machines
ba t algorith