卧式加工中心试验模态分析

应用变时基采样的锤击法对卧式加工中心进行了试验模态分析，研究了该加工中心的动态特性，获得了感兴趣低频段的加工中心低阶固有频率。通过对整机各阶模态振型的分析，找到了加工中心结构的薄弱环节，为进行结构的优化设计指出了改进的方向。     

国内外不锈钢精炼剂研究发展现状

分析了精炼剂组元元素的作用能力，讨论并总结了精炼剂组元的复合作用，介绍了几种国内、国外精炼剂产品，提出发展高效复合精炼剂的想法。     

EFFECT OF γ-IRRADIATION ON CuZnAl SHAPE MEMORY ALLOY

用透射电子显微镜和正电子湮没等分析手段,研究了CuZnAl形状记忆合金的γ辐照效应,此合金经一定剂量的γ辐照后,马氏体转变温度明显提高,而形状记忆功能仍保持不变。     

厚实工作台的铸造工艺

对厚大工作台铸件表面结疤和石墨孔洞缺陷的原因进行了分析，并采取相应措施：采用过滤网挡渣，确保造型、刷涂料、浇注之间有充分的时间间隔，控制涂料的配制和涂刷，控制铁液成分，采用冷铁等，使铸件达到质量要求。     

Web数据挖掘在个性化推荐服务的应用

随着Internet的迅速发展和普及,人们在网络上开展的活动也越来越多,尤其是电子商务的兴起,使得网上购物逐渐成为一种时尚.然而由于互联网上的信息量日益增长,人们面对这些庞大的信息感到无所适从,花费大量时间在搜寻信息上,因此如何为用户提供更好的服务变得非常迫切.本文以数据挖掘理论为基础,分析和探讨了关联规则挖掘算法并应用该方法实现web个性化推荐服务.     

On the model-based networked control for singularly perturbed systems

In this paper, the control of a two-time-scale plant, where the sensor is connected to a linear controller/actuator via a network is addressed. The slow and fast systems of singularly perturbed systems are used to produce an estimate of the plant state behavior between transmission times, by which one can reduce the usage of the network. The approximate solutions of the whole systems are derived and it is shown that the whole systems via the network control are generally asymptotically stable as long as their slow and fast systems are both stable. These results are also extended to the case of network delay.     

线性时滞系统稳定性分析综述

时滞在工程领域广泛存在,对此综述了线性时滞系统的稳定性研究方法.从频域和时域两个角度详细介绍了各种方法的特点,着重讨论基于线性矩阵不等式(LMI)的分析方法,指出保守性是分析的重点.对现有结果的保守性进行比较和评述,并提出了改进的思路.     

Chinese-based research on subject-covered information retrieval supervised by textual semantic domain

By means of analysis of artificial intervention in ready-retrieved text, training set used to compare with new texts from large-scale real texts corpus is provided. It is based on the data-originated presentation of training set that a special formula to calculate semantic cohesion between new texts and training set is devised. The semantic cohesion of new text is the average value of semantic evaluation of all elements involved, and semantic evaluation of an element depends on its semantic relevance with the training set and on the semantic ratio of its domain to synonymous domain. In terms of empirical verification a conclusion is drawn that semantic cohesion is the key measurement standard of textual retrieval. Despite the advantages of textual retrieval, limitations of formula-raised condition and analyst＇s accomplishments make the analysis involved in this paper imperfect.