基于CORBA的无线服务后台软件

无线服务后台软件即C 客户端应用程序，采用C0RBA技术和JaguarCTS三层体系结构开发。三层结构体系中，客户端很小，简单通用，大部分的应用驻留在网络中间服务器上，中间服务器再访问数据库，完成客户端的请求。通过中间件技术平衡网络负载，解决跨操作系统，跨协议等兼客性问题，实现分布式软件集成。在JaguarCTS三层结构基础之上，进一步延伸客户端，灵活构造系统结构。并利用C 面向对象特性打通数据传输通路，建立C0RBAJaguarC 客户应用，实现了无线终端访问网络数据库。     

紧贴保护体钢筋砼老闸墩控爆拆除

介绍了特大体积闸墩爆破拆除时爆破参数的选取和实施方法。新老闸墩一端仅隔4cm，利用钻孔爆破法拆除老闸墩取得了快速，经济，安全的爆破效果。     

尿素对以硝酸铝和葡萄糖为原料合成氮化铝粉末反应过程中相变及反应速率的影响

以硝酸铝(Al(NO3)3·9H2O)和葡萄糖(C6H12O6·H2O)为原料,利用碳热还原法制备氮化铝粉末,研究了尿素对前驱物的制备及前驱物氮化反应的影响,研究发现添加尿素合成的前驱物和未添加尿素合成的前驱物在氮化反应过程中相变和反应速率存在较大差异。在没有添加尿素合成的前驱物的氮化反应过程中,出现了γ-Al2O3、α-Al2O3、AlON和AIN相,该前驱物的反应速率慢,完全氮化需要在1600℃下才能完成。对于添加尿素合成的前驱物而言,在其氮化反应过程中仅出现了γ-Al2O3和AIN相,没有α-Al2O3和AlON的生成,AIN直接由γ-Al2O3氮化生成,该前驱物的氮化反应速率快,氮化反应温度低,在1400℃下即可实现完全氮化。分析讨论了两种前驱物的氮化反应速率不同的主要原因,并利用XRD、SEM等分析方法对粉末进行了表征。     

Stochastic correlative learning algorithms

This paper addresses stochastic correlative learning as the basis for a broadly defined class of statistical learning algorithms known collectively as the algorithm of pattern extraction (ALOPEX) family. Starting with the neurobiologically motivated Hebb's rule, the two conventional forms of the ALOPEX algorithm are derived, followed by a modified variant designed to improve the convergence speed. We next describe two more elaborate versions of the ALOPEX algorithm, which incorporate particle filtering that exemplifies a form of Monte Carlo simulation, to exchange computational complexity for an improved convergence and tracking behavior. In support of the different forms of the ALOPEX algorithm developed herein, we present three different experiments using synthetic and real-life data on binocular fusion of stereo images, on-line prediction, and system identification.     

运用模式分类的雷达抗转发式距离欺骗干扰方法

针对雷达转发式欺骗干扰中的距离欺骗，提出了一种基于模式分类来识别目标与干扰的方法。该方法基于干扰和目标能量和起伏特性差异确定提取特征因子，再利用RBF神经网络方法进行分类。仿真结果表明，该方法具有较好的抗距离欺骗干扰性能。     

工业层析洗梯液回收套用的一种方法

以生产甲胺基阿维菌素苯甲酸盐精品的工艺——工业层析产生的洗梯液回收套用的过程为例，说明利用二元恒沸物的组成随压力的变化而变化的规律，对生产洗梯液进行有效、简洁、合理分离及套用的方法。     

漫步潮汕小故宫

潮汕民间流行一种说法：“朝州厝，皇宫起”只有身临其境，见识了潮汕的世家望族兴建的豪宅庄园，才知此话不含糊。     

马尔柯夫链预测法在证券指数预测中的应用

介绍马尔柯夫链预测法的模糊模型，该模型以集合的模糊划分为依据摒弃了清晰划分的不足．从而保证预测结果的精确性。将该模型应用在证券指数的预测上，取得了绝对误差、相对误差较小的结果。     

基于VHDL实现单精度浮点数的加／减法运算

研究了单精度浮点数加／减法的结构及其设计方法，并在Aldec公司的Active—HDL软件环境下，采用VHDL语言进行设计，并进行了仿真验证，计算精度可以达到10^-7。     

Effect of the graft yield of maleic anhydride on the rheological behaviors,mechanical properties,thermal properties,and free volumes of maleic anhydride grafted high‐density polyethylene

The rheological behavior, thermal properties, and molecular mobility of a series of maleic anhydride (MA) grafted high‐density polyethylenes were characterized and evaluated. The rheological behavior was studied with a Haake minilaboratory. The viscosity of the samples in their melt state decreased with an increase in the graft yield, and this could be attributed to the higher molecular mobility for samples with a higher degree of grafting. The thermal properties were investigated with dynamic mechanical analysis and differential scanning calorimetry. Positron annihilation lifetime measurements were used to study the effect of the degree of grafting on the chemical environment and the atomic‐scale free‐volume properties. It was found that the grafted MA group played a significant chemical inhibition role in positronium formation when the graft yield was low. The results also indicated that the higher the degree of grafting was, the broader the free‐volume distribution was. The relationship between the microstructure and rheological behavior is discussed. © 2007 Wiley Periodicals, Inc. J Appl Polym Sci, 2008