基于支持向量机的缓存预估模型的设计与实现

采用支持向量机(SVM)对网络业务流数据进行预测估计，通过训练样本，从而获得样本以外数据的分布规律。在此基础上，设计了一种网络排队队列缓存的预估模型。实验表明，该模型具有较高的训练效率和很高的估计精确度。     

多分量波场的矢量法叠前深度偏移技术

本文研究了二维各向同性介质中两分量记录的矢量法叠前深度偏移方法———叠前逆时深度偏移方法。首先从弹性波波动方程出发,在交错网格空间中推导了各向同性介质中弹性波逆时延拓的高阶有限差分格式;然后从程函方程出发,采用逆时差分格式求解地下各点的地震波走时。为确保算法满足地震波传播的因果性条件及其对复杂模型的适应性,用扩展波阵面算法追踪波前并搜索全局极小,以上述方法的计算结果作为弹性波逆时偏移的成像条件,实现二维多分量资料的叠前逆时深度偏移。模型试算和实际单炮记录试验表明,叠前逆时深度偏移方法考虑了地震波的矢量特征,是一种有效的矢量波场处理技术。     

凝胶成型法制备加氢催化剂载体的研究与工业化生产

采用凝胶成型法制备加氢催化剂载体，并进行了工业放大试验。通过BET、NH3-TPD、H2-TPR等表征手段及实际的加氢反应性能评价对凝胶成型法制备的载体及加氢催化剂与常规方法进行对比分析，结果表明，采用凝胶成型法所制得的载体及催化剂均具有较高的孔体积和比表面积、较低的堆密度，其机械强度可满足工业要求，总酸量及中强酸酸量得到提升，催化剂活性组分与载体强结合力得到削弱，金属利用率提高，应用性能明显优于传统方法制备的加氢催化剂载体。通过该方法完成了加氢催化剂载体的工业生产，所生产的载体应用于天津院THDS-3加氢精制催化剂取得了良好的工业应用效果。     

线光栅扫描实现衍射图形的矢量化直写

在双光束激光直写系统中,衍射光变图像的光刻是逐像素点进行的.本文提出在该系统中采用狭缝获得光栅线来进行衍射图形的直写,利用一种改进的矢量文件格式,使平台沿着垂直光栅线的方向运动,获得连续的多光点同时直写,直写后的衍射光栅线是连续的,系统的运行效率是逐点光刻的几十倍.编写了与DXF文件的接口直写控制程序,给出了实验结果.     

一种有效的并行汉字／字符串相似检索技术

本文提出了一种有效的并行汉字／字符串相似检索技术．通过引入搜索状态向量及字符一模式匹配向量，该技术将字符串匹配比较转化为简单的整数字位运算，通过对字符串方向相反的搜索有效地实现了多处理机对汉字／字符串的并行相似检索．文中也给出了并行实现算法，同时分析了算法的复杂性．     

中低速率语音波形编码

提出了一种波形编码的新方法，阐述了其编码原理。该编码算法简单，恢复的语音质量较好。由于利用了矢量量化技术，所以该编码速率较低。     

Nitric oxide reduction and carbon monoxide oxidation over carbon-supported copper-chromium catalysts

Carbon supported copper-chromium catalysts are shown to be very active for both the reduction of nitric oxide with carbon monoxide and the oxidation of carbon monoxide with oxygen. Mixed copper-chromium oxide active phases have good activity in the simultaneous removal of nitric oxide and carbon monoxide from exhaust gases. The influence of several catalyst variables has been investigated. The activity per volume of catalyst increases with increasing loading, while the intrinsic activity shows a maximum around C/M=100−50. An optimum catalyst for nitric oxide reduction and carbon monoxide oxidation has a copper/chromium ratio of 2/1. The apparent activation energy for the carbon monoxide oxidation over carbon supported copper-chromium catalysts is 77 kJ/mol, suggesting that the Cu---O bond rupture is the rate-limiting process. The reduction of nitric oxide takes place at higher temperatures. Since all catalysts have a low selectivity for molecular nitrogen formation at lower temperatures, the dissociation of nitric oxide is probably rate determining, resulting in a slightly reduced catalyst system. In an excess of carbon monoxide the reaction is first-order in nitric oxide and zero-order in carbon monoxide. Moisture inhibits the reaction by reversible competitive adsorption, whereas carbon dioxide does not. Oxygen completely inhibits the reduction of nitric oxide due to the more rapid reoxidation of the catalytic sites compared to nitric oxide. Therefore, the reduction of nitric oxide takes place only when all oxygen has been converted and, hence, is shifted to higher temperatures. As a possible consequence, the production of nitrous oxide is reduced. Nitric oxide and molecular oxygen react preferentially with carbon monoxide, so, in an excess of oxidizing component, gasification of the carbon support occurs at higher temperatures after carbon monoxide has been completely consumed.     

无油润滑压缩机活塞环和支承环设计

通过对填充聚四氟乙烯材料活塞环及带卸荷槽结构支承环的研究 ,确定了活塞及活塞环结构 ,建立了带卸荷槽支承环的力学分析模型 ,并将其结果用于结构尺寸的确定。在压缩机工艺指标不变的情况下 ,用该方法设计的活塞环和支承环性能优良。     

深基坑支护结构计算的有限元方法

针对软土地基的特性，本文提出了考虑土体抗力释放与结构模型随施工过程改变的新杆系有限元法，并 编制了求解支护结构内力与变形的计算程序，通过对事例分析．计算结果与实际情况非常吻合，证明了该方法的合理 性。     

Effects of modeling means on properties of monitoring models of spot welding quality

Analyzing and modeling the relation between monitoring information during welding and quality information of the joints is the foundation of monitoring resistance spot welding quality. According to the means of modeling, the known models can be divided into three large categories: single linear regression models, multiple linear regression models and multiple non-linear models. By modeling the relations between dynamic resistance information and welding quality parameters with different means, this paper analyzes effects of modeling means on performances of monitoring models of resistance spot welding quality. From the test results, the following conclusions can be drawn: By comparison with two other kinds of models, artificial neural network (ANN) model can describe non-linear and high coupling relationship between monitoring information and quality information more reasonably, improve performance of monitoring model remarkably, and make the estimated values of welding quality parameters more accurate and reliable.