基于遗传算法的正电子湮没寿命谱拟合  

Fitting positron annihilation lifetime spectroscopy based on genetic algorithm

作　　者：马敏阳[1] 秦秀波[2] 姜小盼[2] 于润升[2] 王宝义[2] 吴伟明[1] 

机构地区：[1]广西大学物理科学与工程技术学院,南宁530004 [2]中国科学院高能物理研究所核分析技术重点实验室,北京100049

出　　处：《核技术》2011年第3期179-182,共4页Nuclear Techniques

基　　金：国家自然科学基金项目(10705031;10835006)资助

摘　　要：正电子湮没谱学是无损探测材料中纳米缺陷的有效手段,通过正电子湮没寿命谱可获得正电子在材料中的湮没状态以及材料内部电子密度分布的信息。本工作基于遗传算法(GA)的全局搜索、随机产生初值的特点,对正电子湮没理论寿命谱进行了多指数函数拟合的尝试,并用MATLAB语言编程实现了曲线拟合,得到的各组分寿命及相应强度。与LifeTime9.0的解谱结果比较,遗传算法求解寿命谱待估参数,具有不用赋经验性初值也不会陷入局部最优的优点。Positron annihilation lifetime spectroscopy （PALS） is a useful technique for nano-defects identification in materials. The positron annihilation status and electrondensity distribution in materials can be obtained by PALS. In this paper, based on characteristics of the genetic algorithm （GA） global search and generation of the initial value randomly, a multi-exponential function fitting of theoretic PALS is attempted, and the curve fitting is performed by using MATLAB. The life expectancy of each component and the corresponding intensity is achieved. The results deduced from the GA and the LifeTime9.0 are compared. The parameters PALS are analyzed by the GA, which does not request assigning experiential initial value and does not involve local optimum.

关 键 词：正电子湮没寿命谱 遗传算法 多指数函数 拟合 

分 类 号：TP311.1[自动化与计算机技术—计算机软件与理论]