Optimization of InAs/GaAs quantum-dot structures and application to 1.3-μm mode-locked laser diodes

The self-assembled growth of InAs/GaAs quantum dots by molecular beam epitaxy is conducted by optimizing several growth parameters, using a one-step interruption method after island formation. The dependence of photoluminescence on areal quantum-dot density is systematically investigated as a function of InAs deposition, growth temperature and arsenic pressure. The results of this investigation along with time-resolved photoluminescence measurements show that the com- bination of a growth temperature of 490℃, with a deposition rate of 0.02 ML/s, under an arsenic pressure of 1×10^-6 Torr （1 Torr = 1.33322×10^2 Pa）, provides the best compromise between high density and the photoluminescence of quantum dot structure, with a radiative lifetime of 780 ps. The applicability of this 5-layer quantum dot structure to high-repetition-rate pulsed lasers is demonstrated with the fabrication and characterization of a monolithic InAs/GaAs quantum-dot passively mode-locked laser operating at nearly 1300 nm. Picosecond pulse generation is achieved from a two-section laser, with a 19.7-GHz repetition rate.     

供热管网中管段阻力系数的辨识方法

提出一种新的管网阻力系数辨识方法,利用管网中少量节点处的压力或流量测量值对所有管段内壁粗糙度进行有效辨识,达到获得管段阻力系数的目的.该方法的辨识模型不仅能灵活利用工程手册所推荐的管道粗糙度等先导信息,还可综合利用多种工况下的辨识值,获得更高精度的辨识结果.基于这种辨识模型,提出一种结合遗传算法(GA)和有效集法(AS)的混合算法求解最优化问题.混合算法改善了局部搜索效率,同时具备较好的全局寻优能力.通过对2个案例管网参数的辨识计算,表明新方法可获得精度较高的辨识结果,并且对较大的管网也能得到稳定的计算结果.     

多组元全动型变焦距物镜高斯光学

设定多个变倍组元和多个补偿组元并应用最优化计算方法,给出了最新型变焦距物镜多组元全动型高斯光学参数的计算方法,它概括了几种所有机械补偿法的变焦方式。     

退化约束条件下的依赖域算法及其收敛性

本文建立带退化线性等式与不等式约束最优化问题的一种依赖域算法，方法用一系列以原点为内点的一般紧集为依赖域。讨论了方法的收敛性，证明了迭代点列必有一聚点为原问题的Ｋｕｈｎ－Ｔｕｃｋｅｒ点，最后，在一定假设下，讨论了算法的超线性收敛性。     

参数估计在最优化模型中的有效性及修正

证明了在涉及随机变量的最优化模型中,如果将参数估计和最优化求解过程分离,实际得到的最优化结果必然低于理论值,即参数估计在最优化模型中总是造成低效的结果。然后给出了两种可行的参数估计修正方法。     

具有不同效用函数的最优投资组合分析

The question of optimal portfolio is that finds the trading strategy satisfying the maximal expected utility function subject to some constraints. There is the optimal trading strategy under the risk neutral probability measure (martingale measure) if and only if there is no-arbitrage opportunity in the market. This paper argues the optimal wealth and the optimal value of expected utility with different utility function.     

GLOBAL COVERGENCE OF THE NON-QUASI-NEWTON METHOD FOR UNCONSTRAINED OPTIMIZATION PROBLEMS

In this paper, the non-quasi-Newton＇s family with inexact line search applied to unconstrained optimization problems is studied. A new update formula for non-quasi-Newton＇s family is proposed. It is proved that the constituted algorithm with either Wolfe-type or Armijotype line search converges globally and Q-superlinearly if the function to be minimized has Lipschitz continuous gradient.