Model Selection for Unsupervised Learning of Visual Context

This study addresses the problem of choosing the most suitable probabilistic model selection criterion for unsupervised learning of visual context of a dynamic scene using mixture models. A rectified Bayesian Information Criterion (BICr) and a Completed Likelihood Akaike’s Information Criterion (CL-AIC) are formulated to estimate the optimal model order (complexity) for a given visual scene. Both criteria are designed to overcome poor model selection by existing popular criteria when the data sample size varies from small to large and the true mixture distribution kernel functions differ from the assumed ones. Extensive experiments on learning visual context for dynamic scene modelling are carried out to demonstrate the effectiveness of BICr and CL-AIC, compared to that of existing popular model selection criteria including BIC, AIC and Integrated Completed Likelihood (ICL). Our study suggests that for learning visual context using a mixture model, BICr is the most appropriate criterion given sparse data, while CL-AIC should be chosen given moderate or large data sample sizes.     

Theoretical and practical implications of creep curve shape analyses for 7010 and 7075

Over tensile stress ranges giving creep lives up to 2000 hours at 373 to 463 K, the creep and creep fracture properties of 7010 are compared with results recorded for 7075 and other precipitation-hardened aluminum alloys. Using the ϕ methodology to quantify the systematic variations in creep curve shape with changing test conditions, the behavior patterns displayed by the different alloys are shown to depend on the extent to which precipitate coarsening causes a progressive loss of creep strength with increasing test duration and temperature. The ϕ relationships also introduce new approaches for rationalization and interpretation of creep data sets, based on the activation energy for lattice diffusion in the alloy matrices and the yield stress or UTS at the creep temperatures.     

变频器调速装置节能浅析

1概述 近十多年来,随着电力电子技术、微电子技术及现代控制理论的发展,变频调速技术已经广泛地用于交流电动机的速度控制,其最主要的特点是具有高效率的驱动性能及良好的控制特性.它是一种高新技术电力节能装置,它通过改变（降低）电源工作频率,来降低动力设备（电机）的转速,减少设备的输出功率,达到输出功率与工作负荷的最佳匹配,实现节能目的,有效地提高了经济效益和产品质量.几乎可以说,有电动机的地方就有变频器,在一切需要进行速度控制的场合,变频器以其操作方便、体积小、控制性能高而获得广泛应用.     

Understanding silica-supported metal catalysts: Pd/silica as a case study

Supported metal catalysts, particularly noble metals supported on SiO₂, have attracted considerable attention due to the importance of the silica–metal interface in heterogeneous catalysis and in electronic device fabrication. Several important issues, e.g., the stability of the metal–oxide interface at working temperatures and pressures, are not well-understood. In this review, the present status of our understanding of the metal–silica interface is reviewed. Recent results of model studies in our laboratories on Pd/SiO₂/Mo(1 1 2) using LEED, AES and STM are reported. In this work, epitaxial, ultrathin, well-ordered SiO₂ films were grown on a Mo(1 1 2) substrate to circumvent complications that frequently arise from the silica–silicon interface present in silica thin films grown on silicon.     

油藏演化的两个极端过程

油气藏形成之后，在地质务件变化不大的情况下，储集层中的油气一般保持相对的稳定。但当地质条件变化较大时，油气藏中则发生明显的蚀变。其中两个极端的演化过程，一是抬升破坏形成生物降解沥青，一是深埋发生原油裂解转化成焦沥青和天然气。生物降解沥青没有一定的形态．在显微镜下呈席状分布，沥青反射率低，抽提物具有明显的生物降解特征。原油裂解形成的焦沥青具有一定的几何形态，沥青反射率很高而无常规抽提物。原油裂解生成的天然气在组分和碳同位素特征上与干酪根裂解生成的天然气有别。图6参15     

Mathematical modelling and numerical simulation of ordered porosity metal materials formation

Ordered porosity metal materials belong to a relatively new class of porous materials named gasars. This paper presents a mathematical model of the complex physical phenomena in the production of gasars. Analyses for heat transfer, solidification kinetics and gas diffusion were coupled to describe the formation of unique gasar structure. Several criterial functions were introduced to provide significant quantitative information about the relationship between the operating technological parameters and the final structure. The computational outcomes of the numerical simulation were compared with the characteristics of real gasar ingots. The model was applied to determine the boundary conditions that would provide approximately constant physical conditions on the solidification front. The structure sensitiveness of gasars with respect to the different technological parameters is discussed.     

气动贯通式潜孔锤反循环连续取心取样钻进新技术在河南钼矿中应用

三道庄钼矿勘探生产中钻遇的地层非常复杂，采用其他方法均无法实现钻进，经过分析比较，决定采用贯通式潜孔锤反循环连续取心（样）钻进新技术来解决该地层钻进难、成孔难、取样难等三大问题。文章从钻进技术特点、工作原理等方面对该技术进行详细分析，得出其具有钻进效率高、岩心（样）采取率高、成孔质量好等突出优点。并介绍了钻进主要设备机具的配套、施工工序、操作要点、遇到问题的解决方法等。在河南钼矿露天采石场应用该项技术，有效避免了塌孔和卡钻事故，工程进展顺利、质量好、岩心（样）采取率在98％以上，取得了良好的效果。针对在生产试验中遇到的某些问题，作者提出了一些改进建议。     

深水海底泥浆举升钻井技术及其应用前景

为了解决深水钻井中遇到的问题，由Conoco公司领导的工业联合项目组研发了海底泥浆举升钻井（以下简称SMD）技术。采用该技术进行深水钻井时，隔水管内充满海水，泥浆用小直径管线从海底返回，在返回环空中形成两个压力梯度，实现双梯度控压钻井。文中着重介绍了SMD的理论依据、系统工作原理、系统关键设备以及SMD井控程序的HAZOP分析等。从技术适用性、经济性、装备以及可能存在的风险等方面对该技术在南中国海钻井中使用的可行性进行研究，并展望其应用前景。     

关于天然气水合物钻采工艺技术进展的研究

目前全球正面临着前所未有的能源危机，天然气水合物由于储量大、能量密度高而将成为一种诱人的新型能源。天然气水合物主要存在于冻土带和海洋，在地下处于低温和高压状态，以现有的技术，钻井和开采都存在很大的困难。简要介绍了天然气水合物的性质及储量，阐述了天然气水合物钻井的六个挑战及目前国外天然气水合物的两种钻探方法；同时阐述了以热力法、降压法和化学试剂注入法为主的天然气水合物开采方法，并针对天然气水合物钻采技术现状指出了下一步的研究方向。     

丙烯腈装置回收塔内聚合物生成原因的分析

针对丙烯腈生产过程中回收塔内聚合物堵塞降液管造成回收塔运行周期短的问题,通过对丙烯腈回收塔内各组分分布的模拟计算,直观显示了回收塔内各组分的浓度分布,发现了回收塔的聚合段与丙烯醛的富集段相吻合的现象,初步得出了回收塔内局部聚合的原因;通过对塔内聚合物的特性分析和评判,进一步证实了“回收塔内局部聚合严重是由于在聚合段丙烯醛聚集引起”这一现象;根据模拟计算结果和实际运行经验,提出了预防聚合物堵塞降液管的措施,并进行了部分生产验证,为丙烯腈生产企业延长装置运行周期提供了理论指导。