高温纵向压印过程的有限元热力耦合分析

利用作者开发的用于模拟二维（包括平面变形和轴对称变形）金属热塑性成形过程的有限元分析系统Ｓ－ＦＯＲＧＥ，对转子钢（２５Ｃｒ２Ｎｉ４ＭｏＶ）方坯高温纵向压印实验过程进行模拟，给出了模拟结果，并与实验结果进行对比，说明了Ｓ－ＦＯＲＧＥ系统的有效性。     

线性系统状态估计器的鲁棒设计方法

在考虑对噪声不确定系统的状态估计器进行鲁棒设计的条件下，提出一种新的简化性能指标函数，据此设计的状态估计器，设计过程为简单直观。     

神经网络的遗传设计及其在甲醛生产建模及优化中的应用

采用遗传算法的机制设计了一种同时对前向网络的连接权值及拓扑结构进行搜索的方法，并采用以隐结点为基本基因型的方式，同时在算法中加入了局部搜索的方法．该算法已被应用于甲醛生产过程的建模及优化，达到了良好的效果．     

不可测变量的预推断控制

将预测控制、推断控制和软测量三者结合起来，以软测量估计器作为预测模型，以一种新的启发式随机搜索方法作为优化算法，构成一种针对不可测变量的新的控制策略──预推断控制．以实际工业装置为背景的仿真研究表明，这种控制策略能够有效地抑制不可测扰动，从而为解决不可测变量的控制问题提供了一条有效途径．     

多媒体声频监控系统开发

研究了把多媒体技术引进监控领域从而改善现有的报警形式．提出了多媒体声频监控系统的结构，并在运用小波分析等数字信号处理算法的基础上，达到了对声音（特别是周期性声音）的变化进行同步声频报警的监控目的．     

拉普拉斯变换在聚合物粘弹性理论中的应用

将拉普拉斯变换应用于聚合物粘弹性理论，使粘弹性材料的特性函数如松弛模量、蠕变柔量以及表示动态力学性能的函数之间的关系简单明了，并用拉氏变换定义松弛谱和推迟谱，将各粘弹函数相互联系起来．     

适于SPICE仿真,拓扑不变的数字电路模型

本文提出了适于ＳＰＩＣＥ程序仿真、拓扑不变的组合逻辑电路模型。采用这种模型，ＳＰＩＣＥ程序就可以直接分析组合逻辑电路，甚至模糊逻辑电路。这种方法扩展了ＳＰＩＣＥ程序的使用范围。     

堆山工程软土地基Biot固结的非线性有限元分析

本文在文献[1]、[2]给出的软土地基Biot固结的有限元构式及程序基础上,对某大型堆山工程分级加荷施工过程中,软土地基的固结变形,附加应力及超孔隙水压力分布,塑性区的开展等进行了弹性非线性有限元的计算分析,得出了在理论和工程应用上一些有意义的结论。本文的计算方法和结果,可供工程技术人员在类似工程的初步设计时借鉴和参考。有些内容尚可进一步研究和探讨。     

Molecular pathogenesis of apolipoprotein E-mediated amyloidosis in late-onset Alzheimer’s disease

Apolipoprotein E (apoE) ɛ4 allele is a genetic risk factor for late-onset familial and sporadic Alzheimer’s disease (AD). In the central nervous system, apoE is secreted mainly by astrocytes as a constituent of high-density lipoproteins. A recent study using apoE knockout mice provided strong evidence that apoE promotes cerebral deposition of amyloid β protein (Aβ). However, no clear explanation of the pathogenesis of apoE-induced AD has been provided. Here we discuss two possible mechanisms by which apoE might enhance Aβ deposition. One is the intracellular pathway in which apoE is internalized by neurons and induces lysosomal accumulation of Aβ and amyloidogenic APP (amyloid precursor protein) fragments, leading to neuronal death. The other is the extracellular pathway in which apoE-containing lipoproteins are trapped by Aβ1–42 deposits mobilizing soluble Aβ peptides and consequently enlarge amyloid plaques. These two mechanisms may operate at different stages of AD pathogenesis and suggest a chaperone-like function for the apoE molecule. Received 4 February 1999; received after revision 9 April 1999; accepted 23 April 1999     

Chemotherapy and immunotherapy of malignant glioma: molecular mechanisms and clinical perspectives

Despite the considerable progress in modern tumor therapy, the prognosis for patients with glioblastoma, the most frequent malignant brain tumor, has not been substantially improved. Although cytoreductive surgery and radiotherapy are the mainstays of treatment for malignant glioma at present, novel cytotoxic drugs and immunotherapeutic approaches hold great promise as effective weapons against these malignancies. Thus, great efforts are being made to enhance antitumoral efficacy by combining various cytotoxic agents, by novel routes of drug administration, or by combining anticancer drugs and immune modulators. Immunotherapeutic approaches include cytotoxic cytokines, targeted antibodies, and vaccination strategies. However, the success of most of these experimental therapies is prevented by the marked molecular resistance of glioma cells to diverse cytotoxic agents or by glioma-associated immunosuppression. One promising experimental strategy to target glioma is the employment of death ligands such as CD95 (Fas/Apo1) ligand or Apo2 ligand (TRAIL). Specific proapoptotic approaches may overcome many of the obvious obstacles to a satisfactory management of malignant brain tumors. Received 8 March 1999; received after revision 27 May 1999; accepted 14 June 1999