面向下一代光通信的VCSEL激光器仿真模型

VCSEL激光器(Vertical Cavity Surface Emitting Laser)在现代光纤通信系统中占有十分重要的地位,为了更好的设计高速VCSEL激光器,基于经典的L-I模型、小信号响应模型,通过拟合实验测量的数据进行模型参数的提取,进而探究了偏置电流、环境温度对于VCSEL激光器L-I特性、频响特性的影响规律.针对经典的L-I模型存在的参数拟合困难、精度不高等缺陷,提出了两种相应的改进模型;针对小信号响应模型只适用于小信号工况下的缺点,引入了一种复合等效电路模型,并对二者进行了比较和分析.     

VCSEL激光器仿真模型在家庭固定网络的应用

探讨了VCSEL激光器工作环境的温度问题和激光器的带宽问题,提出VCSEL激光器的光功率与电流(L-I)模型和器件工作时的电压与电流(U-I)特性曲线模型,利用最小二乘法和高斯牛顿法,对模型初步求解并改进,最终得到最优的激光器工作环境的温度.提出一种基于速率方程的VCSEL的带宽模型(小信号响应模型),并考虑激光器的温度和偏置电流对带宽的影响,利用非线性最优化的方法确定带宽模型,最终得到相应的结论.     

Copula函数的选择：方法与应用

针对目前Copula函数在实际应用中的选择问题,本文通过非参数法得到了它们的分布函数图及其经验分布图并进行了比较,然后利用一种解析法对其进一步的选择,并通过Q-Q图比较了各种模型的拟合程度,最后进行了拟合优度检验,得到了最优的Copula。最后对国内的上证A股指数和上证B股指数进行了实证分析,结果体现了该方法的有效性。     

图像小波域局部统计模型的拟合优度检验

提出了一种图像小波域局部统计模型的拟合优度检验方法, 构造了一个检验的统计量, 其分布为$\chi^2$-分布. 模拟结果表明, 不同分块下的拟合优度检验结果不同, 即从理论上给出了基于小波域局部统计模型的最优分块方法.     

变系数部分线性模型的拟合优度检验

本文考虑变系数部分线性模型的拟合优度检验问题.基于Profile经验似然方法,构造了参数部分和非参数部分的经验似然比检验统计量.并证明了其满足Wilks'现象,进而得到了一定置信水平的拒绝域.最后通过数据模拟,讨论了其检验功效.     

基于COPULA的A、B股信息流动和相关结构分析

金融市场的价格变化体现了市场的信息流动.文章引入copula模型来分析A、B股市场的信息传递.在选择候选copula模型后,利用非参数和极大似然方法得到刻画两个市场信息流动的copula函数的参数,并综合利用AIC检验、KS检验及卡方检验等拟合优度方法判断得到:利用Frank函数来刻画两个市场的相关模式最为合适.这样较灵活的捕捉到了两个市场的信息流动和相关结构.     

基于蒙特卡洛算法的航空兵场站装备广义维修过程研究

研究了适用于航空兵场站可修装备在多级备件、(s-1,s)库存下的广义维修过程解析模型。基于Monte Carlo算法,迭代产生大量样本数据,经过拟合发现该维修过程依分布收敛于一对数正态分布;再针对该样本分别以OLS(最小二乘),ML(极大似然)估计进行参数推断,得到了其稳态分布函数,通过了拟合优度检验。最后解出了该情况下装备的稳态维修度,稳态可用度等参数。对比simlox模型对该装备的评估结果,数据吻合程度较为理想。     

含估计参数的加权经验过程

在讨论(广义)非参数似然比拟合优度检验时,加权经验过程理论是一个非常重要的基础.但对含估计参数的加权经验过程理论,目前文献中很少讨论.对含估计参数的加权经验过程的上界型和积分型两种统计量的近似分布进行了讨论,给出了较一般的结果.所得结论叮以为进一步讨论复合零假设下(广义)非参似然比拟合优度检验提供理论基础.     

A-H双重上市公司股票价格差异研究——基于模糊最小二乘模型

由于中国内地和香港股票市场的市场分割,A-H双重上市公司的股票在两地市场间存在较大的价格差异.本文引入模糊线性回归的概念和方法,在分析A-H股双重上市公司股价差影响因素的基础上,构建模糊线性模型,利用最小二乘法对影响因素进行回归分析.根据拟合优度检验结果,得到影响A-H股价差最关键的因素,并与预期影响进行比较.     

基于蒙特卡洛分析的稳健性优化方法——以白金纳米粒子沉积过程为例

以白金纳米粒子在超临界二氧化碳中的沉积过程的稳健性优化问题为例,首先采用Chrastil模型对白金纳米粒子的沉积过程构建溶解度拟合方程,并进行参数拟合优度分析.然后采用蒙特卡洛仿真方法分析了现有质量约束和成本约束等条件下的沉积过程溶解度最优解,并进行敏感度分析,然后对过程参数进行调整优化,进一步分析了溶解度稳健性最优解,最后提出了基于蒙特卡洛仿真分析的多种约束条件下的过程最优稳健解获取方法.     

基于PLS-SEM的大学生求学满意度研究

在建立大学生求学满意度指数模型的基础上,运用偏最小二乘的结构方程模型对所建立的模型进行了参数求解和检验.实证研究表明:所建立的大学生求学满意度指数模型拟合度较高,并较好地刻画了各变量之间的系数.     

基于改进遗传算法的布局优化子问题

本针对子问题，构造了布局子问题(关于同构布局等价类)的改进遗传算法。将该算法应用于二维布局优化子问题，数值实验表明该算法能够在很好地保持图元的邻接关系的前提下找到子问题的最优解。由于布局优化问题可分解为有限个子问题，所以利用该算法可以找到整个布局优化问题的全局最优解。     

碳纤维及浸润树脂自然老化性能的模型预测

对目前土木工程加固常用的两种碳纤维片材及树脂耐久性进行了对比性试验研究.截至自然老化3年时,已进行了16组共80个试件的试验.考察在我国自然环境条件下抗拉强度的老化.根据复合材料的中值老化剩余强度半经验数学模型,对其进行拟合,回归出CFRP复合材料及其树脂的剩余强度中值方程的相关参数.预测复合材料自然老化剩余强度.对复合材料进行寿命预估,供进一步研究和工程设计参考使用.     

Chaos gray-coded genetic algorithm and its application for pollution source identifications in convection–diffusion equation

In order to reduce the computational amount and improve computational precision for nonlinear optimizations and pollution source identification in convection–diffusion equation, a new algorithm, chaos gray-coded genetic algorithm (CGGA) is proposed, in which initial population are generated by chaos mapping, and new chaos mutation and Hooke–Jeeves evolution operation are used. With the shrinking of searching range, CGGA gradually directs to an optimal result with the excellent individuals obtained by gray-coded genetic algorithm. Its convergence is analyzed. It is very efficient in maintaining the population diversity during the evolution process of gray-coded genetic algorithm. This new algorithm overcomes any Hamming-cliff phenomena existing in other encoding genetic algorithm. Its efficiency is verified by application of 20 nonlinear test functions of 1–20 variables compared with standard binary-coded genetic algorithm and improved genetic algorithm. The position and intensity of pollution source are well found by CGGA. Compared with Gray-coded hybrid-accelerated genetic algorithm and pure random search algorithm, CGGA has rapider convergent speed and higher calculation precision.     

基于“货到人”拣选模式的储位分配问题研究

研究了“货到人”拣选模式下的储位分配问题,以订单拣选过程中搬运货架总时间最短为目标建立了整数非线性规划模型,并证明其为NP-hard问题,分别设计了求解模型的贪婪算法和单亲进化遗传算法。首先根据订单和物品的关联关系对物品进行聚类,基于聚类结果设计了求解模型的贪婪算法。然后设计了直接求解模型的单亲进化遗传算法,遗传算法中采用了0-1矩阵编码、多点基因倒位算子、单点基因突变算子和精英保留等策略,通过合理选取参数,能够很快求解出问题的近似最优解。最后利用模拟算例和一个具体实例进行计算,并对贪婪算法和遗传算法的求解时间和求解效果进行了比较分析。结果显示,对于小规模问题,两种算法均能在较短的时间内以很高的概率得到问题的全局最优解,对于中等规模的实际问题,利用两种算法得到的储位分配方案均优于企业目前采取的基于出库频率的储位分配方案,遗传算法得到的储位分配方案对应的货架搬运次数、货架搬运总时间等均优于贪婪算法。本文设计的遗传算法可以作为智能仓库管理信息系统的核心算法。     

Farm households’ rice production behavior in China under a separability assumption: A metaheuristic optimization approach

The production behavior of farm households has been a topic of interest for a long time. We explore the optimal production strategies from a household's perspective under the assumption of separability between consumption and production. The data used for this study come from farm households located in five provinces of China. To identify the impacts of various parameters on farmers' production behavior, optimal solutions and the households' actual production behaviors are compared. Results show that most of the households overinvested on farm land, labor force, materials, and machinery. Compared to existing production efficiency estimates, households, regardless of size, and exhibit different degrees of nonseparability. The sensitivity analysis shows a similar result, but finds that large households become more efficient if they have a higher production coefficient. The optimized result obtained in this paper not only provides farmers with decision‐support information on efficient resource allocation, but also helps policymakers to formulate better agricultural policies. Recommendations for Resource Managers A household production model is constructed to assess whether production and consumption behaviors are different among Chinese farmers. A genetic algorithm that belongs to a metaheuristic cluster is applied to search for the optimal solution, given the parameters obtained from using a ridge regression model. The following implications could be recommended based on the findings of the paper:

• The goodness of the convergence of the algorithm helps to find the optimal rice production decision, when facing a large number of input variables.
• The evidence that can support the separability among farm household's production behavior which to some degree supports nonseparability.
• As some of the external parameters, such as nonagricultural income, as well as farm scale, are increased, the optimal solution shows the separability among some households.
• Unless farm size is increased substantially, there is a need to support farmers in production behavior so that they can be efficient and profitable rather than just subsistent.

     

求解旅行商问题的基于类Kruskal的混合粒子群算法

本文针对求解旅行商问题的标准粒子群算法所存在的早熟和低效的问题,提出一种基于Greedy Heuristic的初始解与粒子群相结合的混合粒子群算法(SKHPSO)。该算法通过本文给出的类Kruskal算法作为Greedy Heuristic的具体实现手段,产生一个较优的初始可行解,作为粒子群中的一员,然后再用改进的混合粒子群算法进行启发式搜索。SKHPSO的局部搜索借鉴了Lin-Kernighan邻域搜索,而全局搜索结合了遗传算法中的交叉及置换操作。应用该算法对TSPLIB中的典型算例进行了算法测试分析,结果表明:SKHPSO可明显提高求解的质量和效率。     

Parameter optimization and uncertainty analysis in a model of oceanic CO2 uptake using a hybrid algorithm and algorithmic differentiation

Methods and results for parameter optimization and uncertainty analysis for a one-dimensional marine biogeochemical model of NPZD type are presented. The model, developed by Schartau and Oschlies, simulates the distribution of nitrogen, phytoplankton, zooplankton and detritus in a water column and is driven by ocean circulation data. Our aim is to identify parameters and fit the model output to given observational data. For this model, it has been shown that a satisfactory fit could not be obtained, and that parameters with comparable fits can vary significantly. Since these results were obtained by evolutionary algorithms (EA), we used a wider range of optimization methods: A special type of EA (called quantum-EA) with coordinate line search and a quasi-Newton SQP method, where exact gradients were generated by Automatic/Algorithmic Differentiation. Both methods are parallelized and can be viewed as instances of a hybrid, mixed evolutionary and deterministic optimization algorithm that we present in detail. This algorithm provides a flexible and robust tool for parameter identification and model validation. We show how the obtained parameters depend on data sparsity and given data error. We present an uncertainty analysis of the optimized parameters w.r.t. Gaussian perturbed data. We show that the model is well suited for parameter identification if the data are attainable. On the other hand, the result that it cannot be fitted to the real observational data without extension or modification, is confirmed.     

Contrast enhancement for image by WNN and GA combining PSNR with information entropy

A new contrast enhancement algorithm for image is proposed combining genetic algorithm (GA) with wavelet neural network (WNN). In-complete Beta transform (IBT) is used to obtain non-linear gray transform curve so as to enhance global contrast for an image. GA determines optimal gray transform parameters. In order to avoid the expensive time for traditional contrast enhancement algorithms, which search optimal gray transform parameters in the whole parameters space, based on gray distribution of an image, a classification criterion is proposed. Contrast type for original image is determined by the new criterion. Parameters space is, respectively, determined according to different contrast types, which greatly shrink parameters space. Thus searching direction of GA is guided by the new parameter space. Considering the drawback of traditional histogram equalization that it reduces the information and enlarges noise and background blur in the processed image, a synthetic objective function is used as fitness function of GA combining peak signal-noise-ratio (PSNR) and information entropy. In order to calculate IBT in the whole image, WNN is used to approximate the IBT. In order to enhance the local contrast for image, discrete stationary wavelet transform (DSWT) is used to enhance detail in an image. Having implemented DSWT to an image, detail is enhanced by a non-linear operator in three high frequency sub-bands. The coefficients in the low frequency sub-bands are set as zero. Final enhanced image is obtained by adding the global enhanced image with the local enhanced image. Experimental results show that the new algorithm is able to well enhance the global and local contrast for image while keeping the noise and background blur from being greatly enlarged.