基于自回归模型和GA-PNN的海洋平台结构损伤识别研究

文中基于目前神经网络应用的广泛性和对损伤识别的有效性,提出了以自回归模型系数差为初始特征,并通过基于遗传算法(Genetic Algorithm,GA)优化的概率神经网络(Probabilistic Neural Network,PNN)的识别方法,对海洋平台结构进行损伤位置的识别.数据采用数值模拟中海洋平台支撑杆件在...     

Insulin-like Growth Factor 1 Promotes Cell Proliferation by Downregulation of G-Protein-Coupled Receptor 17 Expression via PI3K/Akt/FoxO1 Signaling in SK-N-SH Cells

Insulin-like growth factor 1 (IGF-1) not only regulates neuronal function and development but also is neuroprotective in the setting of acute ischemic stroke. G-protein-coupled receptor 17 (GPR17) expression in brain tissue serves as an indicator of brain damage. As whether IGF-1 regulates GPR17 expression remains unknown, the aim of this study is to investigate how IGF-1 regulates GPR17 expression in vitro. Human neuroblastoma SK-N-SH cells were used. Lentivirus-mediated short hairpin RNA (shRNA) was constructed to mediate the silencing of FoxO1, while adenoviral vectors were used for its overexpression. Verification of the relevant signaling cascade was performed using a FoxO1 inhibitor (AS1842856), a phosphatidylinositol 3-kinase (PI3K) inhibitor (LY294002), and a GPR17 antagonist (cangrelor). Cell proliferation was analyzed using EdU staining; immunofluorescence staining was used to detect the expression and subcellular localization of FoxO1. Chromatin immunoprecipitation was used to analyze the binding of FoxO1 to the GPR17 promoter in SK-N-SH cells. The expression of FoxO1, GPR17, and protein kinase B (also known as Akt) mRNA and protein as well as the levels of FoxO1 and Akt phosphorylation were investigated in this study. IGF-1 was found to downregulate FoxO1 and GPR17 expression in SK-N-SH cells while promoting cell viability and proliferation. Inhibition of FoxO1 and antagonism of GPR17 were found to play a role similar to that of IGF-1. Silencing of FoxO1 by lentivirus-mediated shRNA resulted in the downregulation of FoxO1 and GPR17 expression. The overexpression of FoxO1 via adenoviral vectors resulted in the upregulation of FoxO1 and GPR17 expression. Blocking of PI3K signaling by LY294002 inhibited the effect of IGF-1 on GPR17 suppression. Results from chromatin immunoprecipitation revealed that IGF-1 promotes FoxO1 nuclear export and reduces FoxO1 binding to the GPR17 promoter in SK-N-SH cells. Here, we conclude that IGF-1 enhances cell viability and proliferation in SK-N-SH cells via the promotion of FoxO1 nuclear export and reduction of FoxO1 binding to the GPR17 promoter via PI3K/Akt signaling. Our findings suggest that the enhancement of IGF-1 signaling to antagonize GPR17 serves as a potential therapeutic strategy in the management of acute ischemic stroke.     

深层硬目标易损性的数学描述

给出了深层硬目标效能及其易损性的数学定义,表达了深层硬目标易损性研究的一种思路．通过分析侵彻爆炸毁伤作用,构造了目标完成任务能力表现水平的损失水平函数,函数形式比较简单,其数学意义与毁伤的物理现象吻合较好,通过函数分析能了解深层硬目标在侵爆综合作用下易损性的变化趋势,方便描述目标的毁伤过程．     

混凝土损伤本构模型研究评述

现有混凝土本构关系主要是基于成熟的经典弹塑性模型所建立的,弹塑性模型在数学上较严格,但是与混凝土材料破坏机理不协调,各国学者针对混凝土这类特殊多相复合材料提出了很多基于不可逆热力学理论的损伤本构模型。系统综述了混凝土损伤本构研究的成果,在分析了各个有代表意义的混凝土损伤本构模型基础之上,对比研究了各个模型的特点及各自适用范围,通过总结前人成果,为损伤本构模型研究提供了思路。     

船舶零航速减摇鳍升力仿真研究

应用船舶减摇鳍在零航速时进行减摇,与传统减摇理论有着本质的不同。为了研究船舶零航速减摇鳍在零航速时产生升力的过程,以Weis- Fogh理论为基础,建立了零航速减摇鳍升力模型,对哈尔滨工程大学研制的NJ5型减摇鳍升力模型进行了仿真研究,并应用计算流体力学软件,针对NJ5型减摇鳍进行了数值计算,结果表明数值计算与仿真结果一致,为以后进行水动力试验以及控制规律的进一步研究打下了基础。     

双级装药膛内二次发射内弹道影响因素研究

提出一种采用双级装药进行膛内二次发射以提高弹丸初速的新技术,为提高大口径自动武器的初速提供了一条新思路,并建立了双级装药弹膛内二次发射内弹道模型,揭示了装填条件中弹丸质量配比、装药质量配比、延迟时间和预紧力约束等变化时对膛压和初速的影响规律。     

冲压增程炮弹外弹道理论分析及数值仿真

简述了冲压增程炮弹增程原理,建立了进气道、发动机一体化外弹道计算模型,通过数值仿真得到了155 mm增程弹弹道曲线。通过对仿真结果进行分析,得到了射程随部分设计变量变化的关系曲线,为增程弹总体优化设计奠定了理论基础。     

一类钢筋混凝土材料的动态力学性能研究

利用一级轻气炮对钢筋混凝土靶板进行冲击压缩实验研究,采用Lagrange分析方法对实验数据进行处理,得到流场中各力学量的分布规律。实验结果表明在配筋率较低的情况下,钢筋的加入对混凝土的动态压缩性能改善不明显。但是随着配筋比率的提高,钢筋混凝土的动态压缩强度及延性均有较明显的提高。因此认为钢筋混凝土的动态性能是钢筋的强化作用和其引入所带来的混凝土内部的初始损伤共同作用的结果。基于该实验现象和前人的工作,给出了钢筋混凝土的损伤型动态本构关系,理论预测曲线与实验结果吻合良好。     

拓扑互锁易碎复合材料结构研究

设计出一种新型易碎复合材料结构。它由若干子构件按拓扑互锁原理组合成一种可承载的整体结构,结构中各子构件接触面无粘结,当约束被解除时整个结构即可散落,完成触发碎裂的工作机制。采用高精度模压成型制备出了直角三角板形的拓扑互锁复合材料结构,通过试验研究了该结构在不同横向力作用下的力学行为及其承载能力,并采用有限元法对其进行数值分析。试验结果表明：拓扑互锁易碎复合材料结构在垂直于结构平面的集中载荷作用下,力与位移呈非线性变化。有限元分析结果与试验结果吻合。     

Recent Developments in Predicting Impact and Shock Sensitivities of Energetic Materials （英）

Empirical, quantum mechanical and artificial neural network methods are three usual methods in recent years that were used to predict sensitivity of different classes of high explosives. Some recent developments in predicting sensitivity by various methods are reviewed and discussed for various classes of energetic materials.