铝板纵波电磁超声检测中电磁铁的遗传算法优化设计  

Genetic Algorithms of Electromagnet in Longitudinal Wave Electromagnetic Utrasonic Dtection for Auminum Pate

作　　者：谭良辰 吴运新[1,2,3] 石文泽[1,2] 龚海[1,2] 范吉志 

机构地区：[1]中南大学高性能复杂制造国家重点实验室,长沙410083 [2]中南大学机电工程学院,长沙410083 [3]中南大学有色金属先进结构材料与制造协同创新中心,长沙410083

出　　处：《传感技术学报》2016年第6期813-820,共8页Chinese Journal of Sensors and Actuators

基　　金：国家科技支撑计划基金项目(2014BAF12B01);高性能复杂制造国家重点实验室自主研究课题基金项目(zzyjkt2013-06B)

摘　　要：提出一种提升电磁超声检测能力的电磁铁的遗传算法优化方法,首先采用单目标优化方法,分别以磁感应强度和均匀度为目标,利用响应曲面法(RSM),确定磁感应强度与磁芯,线圈等因素的二阶响应模型,根据响应模型得到优化参数。然后采用多目标优化方法,以磁感应强度和均匀度为目标,采用Matlab与Comsol联合仿真,并结合遗传算法进行优化。两组优化结果表明:多目标优化方法得到的有效区域磁感应强度更高,均匀度更好,通过实验证明优化后纵波信号提升60%,证明优化方法有效,可以将该方法运用到其他形式的电磁铁设计。An original electromagnet genetic algorithm method for improving the detectability of ElectromagneticAcoustic Transducer（EMAT）is introduced. Firstly,Magnetic induction and uniformity were the targets for the sin-gle-objective optimization respectively. The response surface methodology was employed to analysis the importanceof magnetic core,coil on magnetic induction and uniformity,and on top of that,built a second-order response modelto get optimum parameter groups. Secondly,Multi-objective optimization was adapted further considering the unifor-mity of magnetic field,it was optimized by co-simulation of Matlab and Comsol software using genetic algorithm,theconsequence based on multi-objective optimization comparing with RSM was more accurate and it was significantlyincreased comparing with the initial value. At last,the effectiveness of the optimization method was proved by theexperiments,and the experimental longitudinal wave signal rise by 60%,so the method can be applied to the otherforms of electromagnet.

关 键 词：电磁超声换能器 电磁铁 响应曲面法 多目标优化 遗传算法 

分 类 号：TB552[理学—物理]