并行混沌遗传算法在量子级联激光器模型参数优化中的应用  

Parallel Chaotic Genetic Algorithm Assisted Performance Optimization for Quantum Cascade Laser

作　　者：吕默[1] 陈晨[2] 王一丁[1] 

机构地区：[1]吉林大学集成光电子学国家重点联合实验室,电子科学与工程学院,长春130012 [2]吉林大学地球信息探测仪器教育部重点实验室,仪器科学与电气工程学院,长春130061

出　　处：《激光杂志》2016年第3期12-15,共4页Laser Journal

基　　金：中国国家科技支撑计划“煤矿突水、火灾等重大事故防治关键技术与装备研发”重大课题,煤矿用红外CO检测仪(传感器)研发(项目编号:2013BAK06B04)

摘　　要：量子级联激光器具备发射线宽窄,光功率密度高,单色性好与准直度高等特点。随着量子级联激光器技术的日趋成熟,其光谱范围不但覆盖大气透射的三个重要红外窗口,而且已经延伸至太赫兹频段,在化学和生物传感、成像和特殊通信等领域有重要的应用。本文首先从量子级联激光器基本结构和能级模型出发,通过构建量子力学模型实现对量子级联激光器物理特性的仿真,在此基础上引入了并行混沌遗传算法对量子级联激光器有源区设计结构中的关键参数进行了寻优。本文还测试并给出了并行计算的开销时间、加速比性能与收敛性。Quantum cascade lasers have very narrow emission line width, high optical power density, good mono- chromaticity and high collimation. With the quantum cascade laser manufacturing technology matures, it will not only cover the spectral range of three important infrared atmospheric transmission windows, and has been extended to the terahertz band. Quantum cascade lasers have important applications in chemical and biological sensing, imaging, spe- cial communications and other fields. The present paper gives a broad review of the basic structure of QCL and energy level model, then by constructing quantum mechanical model to achieve the simulation of QCL＇ s physical characteris- tics. Furthermore, the parallel chaotic genetic algorithm was adopted to optimize the key parameters of the QCL active region structure. This paper also discusses how to implement parallel code-refactoring in a complex physical model, and tested the overhead time of parallel computing, parallel speedup performance and convergence of the parallel cha- otic genetic algorithm.

关 键 词：量子级联激光器物理建模 并行混沌遗传算法 有源区结构参数优化 并行算法性能分析 

分 类 号：TN247[电子电信—物理电子学]