基于云计算的物联网数据挖掘

物联网是一个巨大的、分布广泛的物与物相连的网络,其上产生的海量数据通常是与时间和空间相关的,具有动态、异构、分布的特性,因此对这些数据的挖掘非常困难,而且耗费时间和内存,效率低下。为了解决这些问题,提出了一种基于云计算的物联网数据挖掘系统:将这些海量数据转化为PML数据文件,并存储在HDFS中,同时把挖掘任务分配到多台节点服务器上并行处理。采用Hadoop平台,将关联规则Apriori算法Map/Reduce化,提高了挖掘的效率。另外,采用副本策略将计算向存储迁移,将失效节点的计算迁移到副本数据存储节点就地执行,降低了数据传输的时间,大为提高了挖掘的效率。     

Synthesis and xylene isomers adsorption of silicalite-1 crystals with intracrystalline mesopores aided by 3-glycidoxypropyl-trimethoxysilane and polyethene glycol

Journal of Porous Materials - Well-defined silicalite-1 crystals with intracrystalline mesopores were successfully synthesized in the presence of 3-glycidoxypropyl-trimethoxysilane (GPTMS) and...     

聚酰亚胺/二氧化硅杂化膜的制备与介电性能的研究

采用溶胶-凝胶法制备了BTDA-ODA聚酰亚胺/SiO2杂化膜,利用红外分光光度计（FTIR）、热重分析仪（TGA）和透射电镜（TEM）研究了杂化膜的微观结构和热性能,并对杂化膜的介电常数（e）和介电损耗（tand）随SiO2粒子含量和电场频率的变化进行了分析和讨论。结果表明：杂化膜的介电常数和介电损耗随SiO2粒子含量的增加而增大,随电场频率的升高而逐渐降低,用考虑到粒子的形状因素和两相间相互作用的EMT模型可以预测聚酰亚胺/SiO2杂化膜的介电常数。     

Simulating the Evolution of Market Shares: The Effects of Customer Learning and Local Network Externalities

This paper develops a simulation model to investigate competitive technology diffusion, and focuses on examining the influence of customer learning on the evolution of market share when both technologies exhibit local network externalities. Results show that: (1) the equilibrium market share of the new technology is determined by two key factors: the characteristics of customer learning behavior and the strength of local network externalities; (2) moderate network externalities can be beneficial for new technology to dominate the entire market when customers adopt belief-based learning rule; (3) moderate learning rate would facilitate the diffusion of new technology when customers make their decisions based on reinforcement learning; (4) the decay of customer learning and the proportions of imitators in market would help the old technology establish advantage by maintaining demand inertia of customer. The joint effects of psychology of customer behavior and local interactions of customers offer a new mechanism to explain the diffusion of technology in a competitive market with network externalities.     

Toluene oxidation on Pd catalysts supported by CeO2–Y2O3 washcoated cordierite honeycomb

A novel CeO₂–Y₂O₃ (CY) washcoat on cordierite honeycomb was prepared by an impregnation method, which was used as a support to prepare a Pd catalyst. A model reaction of the complete combustion of toluene was conducted to evaluate the performance of the developed Pd/CY catalyst. The CY washcoat support and the Pd/CY catalyst were characterized by XRD, Raman spectroscopy, H₂-TPR and SEM techniques. The results show that compared with conventional washcoat the CY washcoat has better adhesion and higher vibration- and heat-resistance. The CY washcoat can anchor well Pd onto the cordierite honeycomb substrate. The formation of a CeO₂–Y₂O₃ solid solution and the steady present of PdO occur at high calcination temperatures, resulting in a better thermal stability. On a Pd/CY catalyst calcined at 500 °C, a 99% of toluene conversion was obtained at 210 °C, and it was stable for reaction time up to 30 h.     

Fe（Ⅲ）-5-Br-PADAP络合物光化学还原反应的研究

研究了溶液酸度、光强度、有机酸及其浓度对Fe(Ⅲ)-5-Br-PADAP络合物光化学还原反应的影响,并对该光化学还原反应的机理作了初步探讨.结果表明,在pH为5.5的HAc-NaAc缓冲溶液中,Fe(Ⅲ)-5-Br-PADAP络合物的光化学还原反应速度较快;光强度增加,该光化学还原反应速度加快;甲酸等有机酸能加速该光化学还原反应;乙二酸、丙二酸能明显抑制该光化学还原反应;能促进反应的有机酸浓度增加,该光化学还原反应的速度加快.     

汽车虚拟仪表盘的仿真设计

汽车仪表盘反映了汽车行驶过程的多种参数,设计虚拟仪表盘仿真系统在教学科研中具有重要意义。在分析汽车仪表盘功能的基础上,提出了汽车虚拟仪表盘总体设计方案。编写通信程序实现了上位 PC 机和下位单片机系统的数据采集发布功能。应用VB开发了数字化汽车虚拟仪表盘界面与性能仿真系统,可模拟发动机转速、车速、温度、时间及里程等信息实时显示在计算机界面上;操作控制按钮可实现汽车转向灯指示功能。经实际试验运行显示该仿真系统工作正常,实现了对汽车仪表盘的仿真效果。     

一类结构不确定非线性系统的鲁棒镇定

基于鲁棒控制Lyapunov函数,首先给出一个函数为鲁棒控制Lyapunov函数的等价条件,进而获得该类系统可鲁棒稳定、可鲁棒实用稳定、可鲁棒渐近稳定的充分条件,并构造了与之相应的逐点最小范数的连续状态反馈控制律.仿真实例说明了所采用方法的有效性.     

浙江高等教育均衡发展的思考

浙江高等教育均衡发展,主要是指浙江省内不同地市之间、城乡之间、同一地市不同高校之间、同一高校不同群体之间的均衡发展。本文主要研究浙江高等教育的“三大均衡”,即：地市间实现均衡发展,只研究11个地市高校的均衡布局问题;城乡间实现均衡发展,只研究城乡高校的均衡布局,尤其是县级社区学院的发展问题;高校间实现均衡发展,只研究不同高校学科带头人的均衡布局问题。     

捷联式惯性导航系统初始对准扰动补偿控制算法

提出了一种捷联惯性导航系统初始对准扰动补偿二级控制器的设计方法。选取载体的飞行姿态、速度与位置增量建立了扰动补偿控制的状态空间数学模型,基于Backstepping控制方法,通过引入虚拟反馈控制变量设计了非线性控制器,保证大初始扰动条件下系统状态稳定收敛到系统的平衡点集。在平衡点集,设计了H2最优状态反馈控制律,以保证系统状态在平衡点集的渐进收敛。仿真和系统试验结果表明,这种二级补偿控制方法能有效抑制初始对准扰动误差,导航解算精度能够满足工程实际需要。