基于对面红外检测的缺陷混沌-LM混合识别算法  

Chaos-LM Hybrid Algorithm for the Defect Identification With Opposite-Side Thermographic Testing

作　　者：吕事桂[1] 杨立[1] 范春利[1] 孙丰瑞[1] 

机构地区：[1]海军工程大学动力工程学院,湖北武汉430033

出　　处：《工程热物理学报》2013年第7期1352-1356,共5页Journal of Engineering Thermophysics

基　　金：国家自然科学基金资助项目(No.50906099)

摘　　要：在对面红外检测条件下,利用混沌算法全局收敛和LM算法快速收敛的相对优势提出了识别试件内部矩形缺陷的混沌-LM混合算法,并引入了相对敏感系数的概念用于评估缺陷参数的定量可检测性,同时采用附加内置热电偶测温的方法弥补了红外测温信息的不足。通过数值实验证明了以上研究内容的有效性。With the opposite-side thermographic testing,the chaos-LM hybrid algorithm is introduced here to identify the rectangle defect inner specimen,based on the global optimum of chaos optimization algorithm and fast convergence of Levenberg-Marquardt（LM） method.And the relative sensitivity coefficient is brought forward to assess the quantitative inspectability of the defect.Besides, in order to cover the insufficiency of the infrared temperature measurement,some inner temperature is also measured by thermocouple.In this paper,the effectiveness of the above research is certified by the numerical experiments.

关 键 词：红外检测 导热反问题 缺陷识别 混沌算法 Levenberg-Marquardt法 全局最优 

分 类 号：TK38[动力工程及工程热物理—热能工程]