用遗传算法求解多目标0/1背包问题

扼要介绍多目标优化的Pareto最优性概念 ,研究搜索多目标 0 1背包问题Pareto最优解集的快速遗传算法 (FPGA :fastParetogeneticalgorithms) .FPGA采用种群中非支配解的层次评价可行解的适应值 ,提出了一种快速非支配解层次辨识算法 ,辨识算法仅有O(n2 )数量级的计算复杂性 ;采用基于聚类概率排挤的小生态技术维持种群多样度和Pareto最优解集的分布均匀性。对多种多目标 0 1背包问题的仿真优化实验结果表明 ,FPGA能够以有效的计算成本搜索到精度高的、分布均匀的高质量Pareto非劣解集 ,其收敛速度和收敛准确性一致地优于代表性的强度Pareto进化算法 (SPEA) .     

基于变结构的移动机器人侧向控制器的设计

采用基于后推与模糊滑模控制相结合的方法设计了移动机器人的变结构状态反馈控制器,实现了对道路跟踪侧向误差的渐近镇定控制.应用Lyapunov定理推导了移动机器人满足非完整性约束的时变光滑状态反馈镇定控制律;将状态控制区域分为远离奇异点的稳定工作区域与含有奇异点的非稳定区域2部分.设计模糊控制嚣来控制移动机器人的状态从非稳定区域向稳定区域转变,并采用移动机器人Amigobot作为实验平台验证了控制器设计的有效性.研究结果表明:侧向误差较大时,模糊控制器确保移动机器人在稳定区域内沿着模态切换线减小误差;当侧向误差较小时,时变光滑状态反馈控制实现对移动机器人的平稳镇定.在控制器设计中,将移动机器人的平移速度视为渐变参数,根据侧向误差能量函数的变化进行模糊自适应调节.     

椭圆曲线加密算法在移动办公系统中的应用

从移动办公的安全性要求出发,阐述了椭圆曲线及椭圆曲线加密体制的优点,主要研究了在移动办公系统中用椭圆曲线加密算法替代传统的公钥密码算法RSA的实现,并证明了在移动办公系统中椭圆曲线加密算法不仅满足移动办公的安全性要求,而且优于传统的加密算法.     

SEM analysis of the changes of carbon layer structure of mesocarbon microbeads heat-treated at different temperatures

Mesocarbon microbeads (MCMB) with narrow size distribution, excellent sphericity and no obvious conglutination have been prepared with a coal tar pitch containing quinoline insolubles (QI) as the raw material. Optical microscopy and scanning electron microscopy (SEM) are used to examine the structure of the MCMB. It has been found that SEM technique shows the structural information of MCMB clearly in the form of micrographs even when the structure of MCMB is complex, while the optical technique is useful for analysis of the regular structure of the mesophase spheres but cannot be effectively used to analyze either the complex structure of the green MCMB or the structures of the further heat-treated ones at different temperatures. According to the characteristics of the carbon layers, the structures of the as-prepared MCMB in the present experiment could be classified as (I) Parallel Layer type structure and (2) Bent Layer type structure with the carbon layers gathering at one or two points in the MCMB. In the experiments, SEM is also utilized to investigate the structures of MCMB that are heat-treated at different temperatures. It has been found that the MCMB without any further heat-treatment show no layered-carbons, while the ones heat-treated at temperaturehigher than 1000℃ exhibit obvious layered carbons across their sections. When increasing the heat-treatment temperature, the carbon layers become thinner and flatter.     

Effects of Ca on the distribution, structure and morphology of aluminum species in polyaluminum chloride

The flocculation efficiency of polyaluminum chloride (PAC) is closely related to the distribution and structure of its AI species, and Al13 is the optimal species in PAC for flocculation. A series of PAC containing Ca was prepared by adding Ca before and after the basifying reaction. The effects of Ca on the Al species were studied by ^27Al nuclear magnetic resonance (NMR) and atomic force microscope (AFM) techniques. The experimental results show that the introduction of Ca increases the content of Alm and Al13 in PAC and decreases their chemical shifts in NMR spectra due to the electric repulsion between the positive Ca species and AI species and the formation of AI-O-Ca complexes.With the rise of Ca/AI molar ratio, the AI species in PAC tend to scatter. It is observed that the formerly branch-aggregated clusters are tending to form granule-aggregated ones whose diameter gets smaller and smaller, and the floccule aggregates are formed at a higher Ca/Al molar ratio. The introduction of Ca to PAC, which increases the Al13 content, is certainly to enhance the flocculation efficiency of PAC in water treatment.     

Deep-water Earliest Oligocene Glacial Maximum （EOGM） in South Atlantic

The most prominent cooling event of the Earth surface during Cenozoic in the long-term transition from a non-glaciated planet, or “green-house world”, to a polar, glaciated planet, or “ice-house world”, is the Earliest Oligocene Glacial Maximum (EOGM) above the Eocene/Oligocene boundary at about 33.7 Ma. Planktonic and benthic foraminiferal oxygen and carbon isotopes, carbonate content, and coarse fraction, along with high-resolution color reflectance and magnetic susceptibility records during 35-30 Ma, from deep-water Sites 1262 and 1265, Ocean Drilling Program (ODP) Leg 208 in South Atlantic, reveal the global cooling event occurring in both surface and deep oceans. The results show that the earliest Oligocene δ^18O values during 33.5-33.1 Ma represent the magnitude of continental ice sheets on east Antarctica and indicate the large decrease in both surface and deep water temperatures of worldwideoceans. The δ^13C records show the large excursion during the period of EOGM event and indicate some types of shift in global carbon reservoir, probably demonstrating the sudden increase in organic carbon burial rates and the changes in the distribution and timing of production. At the same time, lithologic composition, carbonate content, color reflectance, and coarse fraction brought about significant changes close to the Eocene/Oligocene boundary, reflecting the abrupt deepening in the carbonate compensation depth (CCD). Changes in carbonate content were revealed from the color reflectance identify periodicities associated with eccentricity of the Earth‘s orbit (100 and 400 ka), further indicating orbitally forced global climate variations in the Early Oligocene.     

高新技术的发展与我国劳动就业的对策

高新技术与劳动就业是我国社会经济发展的两大战略问题。高新技术的发明和应用促进着经济的发展、改变着社会的生活 ,从而影响与经济和社会生活广泛联系的劳动就业。本文从高新技术的广泛影响出发 ,深入分析高新技术的发展对我国就业结构和就业形势的影响 ,并针对高新技术的发展引起的结构性失业现象提出我国应该采取的有效对策。     

集装箱多式联运系统装卸与运输一体化优化问题

集装箱多式联运（Container Multimodal Transport）是一种以实现货物整体运输的最优化效益为目标的联运组织形式。本文针对集装箱多式联运系统优化的一个核心问题——装卸与运输一体化作业优化问题，进行深入分析，提出一个分析模型，为该模型开发了实用优化算法，并通过与基准问题（benchmark problems）的对比实验显示了该算法实用性及有效性。     

重置变结构神经网络及其在风险投资项目评估中的应用

基于神经网络的结构直接影响到网络性能的优劣,进而影响其推广使用,尝试将重置算法应用于经典BP神经网络的结构优化,研究了重置算法中最佳重置时间的性质,提出一种重置变结构经典BP神经网络。通过实例,将重置变结构经典BP神经网络应用于风险投资的项目风险评估中,并与经典BP神经网络的实验结果进行对比。结果表明,重置算法的引入有效地解决了神经网络结构的优化问题。     

网络--探索复杂性的新途径

简单系统由于非线性关系可以呈现复杂行为，而复杂系统可以遵循简单规则自组形成、近年来，作为探索复杂系统结构和功能的复杂网络引起了国际科学界的高度重视，已成为数理科学、生命科学、社会科学、技术科学和管理科学的研究热点之一．论文首先介绍了在复杂网络研究方面已经取得的某些进展，从典型模型．重要概念，网络分类一直到网络结构涌现，其中也包括了作者所做的工作；然后对某些值得注意的问题做了简短的评述．