虚拟环境下机床主传动系统故障诊断的仿真实现

本文介绍了用AutoCAD完成机床主传动系统三维实体造型，用VisualC 、Visual LISP等工具完成主轴箱轴承、齿轮、轴等部位在运行中出现故障时的立体动画。经压电晶体加速度传感器实时检测机床的振动信号，对这些信号进行分析(统计处理、频谱分析等)提取出故障特征，经人工神经网络、模糊决策及专家系统等手段进行决策，判定故障部位和程度，可在微机上直观地观察主轴箱在运行过程中，各故障具体部位及损伤程度的虚拟现实情况。     

基于危险理论的人工免疫原理与应用*

侧重以危险理论的基本机制为线索,对危险理论的研究现状加以系统论述.首先以危险理论的免疫模型入手,归纳提炼出仿真机理并建立起隐喻算法;介绍其中具有代表性领域中的应用情况;最后对基于危险理论的人工免疫研究的下一步工作进行展望.     

重力坝弹性参数反演的灵敏度分析及一种新的反演算法

依据结构灵敏度分析理论,采用基于有限单元法的解析灵敏度分析方法,对影响重力坝材料参数反演质量的位移观测点位置进行了研究,得到了坝顶位移最适合于反演大坝弹性模量的结论。同时引入一种新的群集智能优化方法—人工鱼群算法（AFSA）,根据大坝坝顶的增量位移进行混凝土大坝和坝基岩石材料参数的反演。算例分析表明：AFSA能够稳定快速地收敛到所求参数的全局最优值;同时加入不同水平的噪音对该算法的数值稳定性进行了研究,结果表明算法稳定性良好。     

Continuous size fractionation of magnetic nanoparticles by using simulated moving bed chromatography

The size fractionation of magnetic nanoparticles is a technical problem, which until today can only be solved with great effort. Nevertheless, there is an important demand for nanoparticles with sharp size distributions, for example for medical technology or sensor technology. Using magnetic chromatography, we show a promising method for fractionation of magnetic nanoparticles with respect to their size and/or magnetic properties. This was achieved by passing magnetic nanoparticles through a packed bed of fine steel spheres with which they interact magnetically because single domain ferro-/ferrimagnetic nanoparticles show a spontaneous magnetization. Since the strength of this interaction is related to particle size, the principle is suitable for size fractionation. This concept was transferred into a continuous process in this work using a so-called simulated moving bed chromatography. Applying a suspension of magnetic nanoparticles within a size range from 20 to 120 nm, the process showed a separation sharpness of up to 0.52 with recovery rates of 100%. The continuous feed stream of magnetic nanoparticles could be fractionated with a space-time-yield of up to 5 mg/(L∙min). Due to the easy scalability of continuous chromatography, the process is a promising approach for the efficient fractionation of industrially relevant amounts of magnetic nanoparticles.     

Simulated Microgravity Inhibits the Proliferation of Chang Liver Cells by Attenuation of the Major Cell Cycle Regulators and Cytoskeletal Proteins

Simulated microgravity (SMG) induced the changes in cell proliferation and cytoskeleton organization, which plays an important factor in various cellular processes. The inhibition in cell cycle progression has been considered to be one of the main causes of proliferation inhibition in cells under SMG, but their mechanisms are still not fully understood. This study aimed to evaluate the effects of SMG on the proliferative ability and cytoskeleton changes of Chang Liver Cells (CCL-13). CCL-13 cells were induced SMG by 3D clinostat for 72 h, while the control group were treated in normal gravity at the same time. The results showed that SMG reduced CCL-13 cell proliferation by an increase in the number of CCL-13 cells in G0/G1 phase. This cell cycle phase arrest of CCL-13 cells was due to a downregulation of cell cycle-related proteins, such as cyclin A1 and A2, cyclin D1, and cyclin-dependent kinase 6 (Cdk6). SMG-exposed CCL-13 cells also exhibited a downregulation of α-tubulin 3 and β-actin which induced the cytoskeleton reorganization. These results suggested that the inhibited proliferation of SMG-exposed CCL-13 cells could be associate with the attenuation of major cell cycle regulators and main cytoskeletal proteins.     

CO2腐蚀的神经网络分析

利用神经网络的分析方法 ,对CO2 腐蚀进行分析建模 ,从系统的观点出发 ,建立了碳钢含碳量、CO2 分压、温度、流速与腐蚀速率之间的非线性对应关系 ,为准确的数学建模、预测设备的使用寿命和相关专家系统的研制开拓了新的思路     

焊接专家系统设计及开发技术

论述了焊接专家系统的基本原理和结构,着重介绍了开发焊接专家系统的技术,包括传统的开发工具和新的先进技术在焊接专家系统中的应用.     

Ti5Nb人工钉扎中心NbTi超导体的磁通钉扎机制分析

采用棒基组合的装配方法，用热挤压和冷加工相结合的工艺制备了岛状Ti5Nb人工钉扎中心NbTi超导体，其中Ti5Nb体积含量为10％。通过对此超导体进行详细的磁通钉扎机制分析，认为样品的磁通钉扎力特征主要由正常相钉扎和Δκ扎相结合的机制决定。根据分析的结果，采用多源标度方法，很好地拟合了不同线径样品的磁通钉扎力密度随磁场变化的曲线。拟合的结果表明：样品在较低场下的性能主要由正常相钉扎机制决定，而在较高场下的性能主要由Δκ钉扎机制决定。并且随着线径的减小，正常相钉扎力密度和Δκ钉扎力密度均有不同程度的提高。     

基于人工智能的钛合金热变形工艺参数优化

在深入分析热变形工艺参数对Ti-15-3合金显微组织及成形载荷的影响的基础上,以变形温度、变形程度和变形速率等热变形工艺参数作为设计变量,以显微组织和成形力的最佳综合为目标,建立了该合金热塑性成形工艺参数的多目标优化数学模型。以显微组织参数和成形力的人工神经网络预测模型作为优化算法的知识源,将人工神经网络与修正的遗传算法相结合,对Ti-15-3合金的热塑性成形工艺参数进行优化。结果表明,提出的修正的遗传算法是有效的,采用将其与人工神经网络相结合的方法对钛合金的热塑性成形工艺参数进行优化是可行的。