考虑多因素交互作用影响的权重确定

引入岩石工程系统（RES）理论的交互作用矩阵构造和编码原理,提出了一种新的考虑多因素交互作刷影响的权重确定方法。基于改进BP网络推导了新的权重确定方法的算法流程,其网络输入层与隐层间以及各隐层间变换函数采用双曲正切函数S型函数tansig;f（x）=2／（1＋exp（-2x））-1,隐层与输出层间采用logsig函数：f（x）=1／（1＋exp（-x）,并采用动量法和自适应学习率相结合训练网络。通过VCR法采场岩爆实例分析,得出了交互作用下岩爆倾向性8个影响因素的作用权重大小,展示了该法的实际应用。     

From immune response to cancer: a spot on the low molecular weight protein tyrosine phosphatase

Reversible tyrosine phosphorylation is a key posttranslational regulatory modification of proteins in all eukaryotic cells in normal and pathological processes. Recently a pivotal janus-faced biological role of the low molecular weight protein tyrosine phosphatase (LMWPTP) has become clear. On the one hand this enzyme is important in facilitating appropriate immune responses towards infectious agents, on the other hand it mediates exaggerated inflammatory responses toward innocuous stimuli. The evidence that LMWPTP plays a role in oncological processes has added a promising novel angle. In this review we shall focus on the regulation of LMWPTP enzymatic activity of signaling pathways of different immunological cells, the relation between genetic polymorphism of LMWPTP and predisposition to some type of inflammatory disorders and the contribution of this enzyme to cancer cell onset, growth and migration. Therefore, the LMWPTP is an interesting target for pharmacological intervention, thus modifying both inappropriate cellular immune responses and cancer cell aggressiveness. Received 15 August 2008; received after revision 06 October 2008; accepted 14 October 2008     

AHP逆序研究

深刻分析了ＡＨＰ中逆序产生的原因,指出文献［２］中给出的保序方法的缺陷,提出了新的ＡＨＰ逆序的解决方案     

基于可变模糊云的既有玻璃墙安全预警

为了科学有效地进行既有玻璃幕墙的安全预警,考虑指标权重的不确定性以及指标出现极端数值的情况,提出了一种基于组合权重与可变模糊云的既有玻璃幕墙二维预警方法。首先,以实际工程为背景,通过安全风险因素的识别,构建了一套定性与定量指标相结合的既有玻璃幕墙安全预警体系;随后,基于可变模糊集理论改进了逆向云发生器,提出了可变模糊云算法并完成了指标预警云的生成;接着,在正态云与熵权法的基础上,提出了信息云组合权重法,继而实现了权重的组合;最后,通过虚拟云与模糊熵的计算,分析了既有玻璃幕墙的安全状态,实现了既有玻璃幕墙的二维预警与警情状态的可视化。研究结果表明：与其他预警方法相比,该方法不仅实现了警情的基本判断,还能提出更为有效的安全处理意见,从而为既有玻璃幕墙的安全管理提供参考。     

Population characteristics of four deep‐water pandalid shrimps (Decapoda: Caridea) in the northern Aegean Sea (NE Mediterranean)

The population characteristics of four deep‐water pandalid shrimps (Decapoda: Caridea) were studied in the northern Aegean Sea in samples collected annually (1997–2000) in summer, through experimental trawl, at depths down to 600 m. Females were larger than males in Plesionika martia and Plesionika heterocarpus; no significant differences in size of males and females were found in Chlorotocus crassicornis and Plesionika antigai. A trend of increasing size with depth was observed, as well as a linear increase in the proportion of females with increasing size. Sex ratio was skewed towards females at all depths in P. antigai, while in P. heterocarpus and C. crassicornis males dominated the intermediate depths. Negative allometries were apparent in the carapace length–weight relationships in males of all species. Isometric relationships were found in females apart from in P. heterocarpus. The population patterns conform to relevant Mediterranean reports. However, some population differences exist that may reflect the distinct environmental conditions prevailing in the study area.     

主用户状态改变和低信噪比环境下的weight-p能量检测算法

在认知无线电网络中,主用户状态改变和低信噪比都会造成频谱检测的性能下降.本文提出了一种新的加权(weight-p)能量检测算法,用于抵抗主用户状态改变和低信噪比对认知用户检测性能的影响.为减少实现复杂性和节约需要的功耗,我们将weight-p能量检测器的最优权值建模成一个最小采样时间(MST)的优化问题,找出了最优权值和次优权值.仿真表明,在主用户状态改变和低信噪比的场景下,本文提出的weight-p能量检测算法可以提高认知用户的检测性能和降低虚警概率,并且在获得相同检测性能的前提下可以压缩检测时间.     

价格关联模型和价格体系改革

本文用投入—产出原理研究价格体系和价格体系改革,建立一个价格关联模型用以描述重要经济变量与价格之间的联系,用模糊聚类的方法对国民经济的主要变量作分类.价格改革的难点是:为防止由工资或原料成本推动的通货膨胀,保持当前的生活水平,使新的利税率向量与目标利税率向量的差别最小,哪一类部门应提价,价格和工资的增幅该多大?混合使用本文模型和常用统计方法,难题得以解决.     

基于层次分析法的油液诊断特征属性的选择

利用油液监测信息进行故障诊断,由于没有区分监测手段及其特征属性的优势,往往使得通过不同监测手段获得的信息所得到的结论可能不同,甚至矛盾.采用具有这方面独特优势(如整合定性定量信息等)的层次分析法(AHP),通过奇异值分解判断矩阵,成功地为大型冲压自动生产线状态监测解决了油液监测手段及其特征属性的权重优势排序和选择问题,使状态监测和故障诊断更有针对性,从而达到增强诊断力的目的.     

基于属性坐标的变权加权和模型

讨论了在加权和方法中引入变权的必要性和合理性.以决策变量之和作为变权向量的自变量,利用属性坐标学习获得变权向量,建立了一种基于属性坐标学习和分析的变权的加权和决策模型,并给出了应用的例子.     

基于精密称重传感器的易开盖拉力测量分析仪的研制

为满足我国食品饮料容器金属包装行业的迫切需要,研制出了易开盖拉力测量分析仪.该仪器采用精密称重传感器作为测力敏感元件,单相可逆电动机作动力,通过USB接口与笔记本电脑(PC机)配合使用,操作界面友好.具有测试曲线显示、打印和数据存储、分析功能,适用于各类规格的易开盖拉力测量和分析.该仪器精度≤0.01N(牛顿),高于国家标准两个数量级.