Hyperparameter on-line learning of stochastic resonance based threshold networks

Aiming at training the feed-forward threshold neural network consisting of nondifferentiable activation functions, the approach of noise injection forms a stochastic resonance based threshold network that can be optimized by various gradient-based optimizers. The introduction of injected noise extends the noise level into the parameter space of the designed threshold network, but leads to a highly non-convex optimization landscape of the loss function. Thus, the hyperparameter on-line learning procedure with respective to network weights and noise levels becomes of challenge. It is shown that the Adam optimizer, as an adaptive variant of stochastic gradient descent, manifests its superior learning ability in training the stochastic resonance based threshold network effectively. Experimental results demonstrate the significant improvement of performance of the designed threshold network trained by the Adam optimizer for function approximation and image classification.     

环境规制强度影响能源效率的门槛效应研究——基于PSTR模型

利用非径向-SBM和CCR模型测算了我国30个省市(暂不包括西藏和港澳台地区)2007-2016年间的环境规制强度以及能源效率,构建面板平滑转移回归模型(PSTR),分析了环境规制强度对我国能源效率的连续非线性变化的影响.研究结果显示,全国环境规制强度对我国能源效率存在非线性效应,二者之间呈倒“U”型关系.环境规制强度小于门槛值0.6376时,环境规制对能源效率产生创新补偿效应,当环境规制强度高于门槛值0.6376时,环境规制对能源效率产生遵循成本效应.而我国东西部地区的环境规制强度与地区能源效率间却存在着与之相反的“U”型效应.     

The energy-critical nonlinear wave equation with an inverse-square potential

We study the energy-critical nonlinear wave equation in the presence of an inverse-square potential in dimensions three and four. In the defocussing case, we prove that arbitrary initial data in the energy space lead to global solutions that scatter. In the focusing case, we prove scattering below the ground state threshold.     

航空导管弯曲试验的疲劳裂纹萌生寿命

飞行器液压导管受接头和卡箍等约束，在使用的振动环境中，会因弯曲应力而导致破裂，影响到飞行安全．本文对飞行器液压系统通用的不锈钢导管的裂纹萌生寿命进行了试验研究．首先在对8 mm、12 mm 无缺陷导管和含U 型缺口8 mm 导管的疲劳试验和有限元分析的基础上，得到了导管的最大拉应变-裂纹萌生寿命数据．然后采用基于强度极限和弹性模量估算法的Manson-Coffin 公式来预测导管裂纹萌生寿命．最后引入加载类型修正系数、表面质量修正系数、试样尺寸修正系数、应力集中敏感系数和有效应力集中系数，使修正后的公式对三种类型的导管均有较好的裂纹萌生寿命预测精度．     

激光驱动复合飞片冲击起爆HNS-Ⅳ实验研究

激光驱动飞片冲击起爆技术具有很强的抗电磁干扰能力和可以直接起爆钝感炸药等优点，能够满足现代战场对火工系统的高安全性和高可靠性要求。HNS-Ⅳ是最适合激光驱动飞片冲击起爆技术的药剂。本文中在6种不同激光能量下，测试了Al/Al₂O₃/Al复合飞片和Al单层飞片对HNS-Ⅳ药剂(装药密度为1.5 g/cm3)的冲击起爆情况。实验实现了激光驱动飞片对HNS-Ⅳ的成功起爆。在217~245 mJ激光能量范围内，激光驱动Al/Al₂O₃/Al复合飞片均可成功完全起爆HNS-Ⅳ药柱。Al单层飞片均未成功起爆HNS-Ⅳ药柱。飞片冲击压力对激光驱动飞片冲击起爆HNS-Ⅳ起决定作用。     

BrCN分子的电子碰撞离子对解离

离子对解离是一类重要的分子过程，常发生于分子被激发到超激发态. 与光激发的离子对解离实验研究不同，电子碰撞的相关过程研究尚存在实验挑战，特别是在测定其阈值方面. 本文报道了相关的利用单色化电子碰撞分子的实验研究进展. 以BrCN→Br^-+CN⁺离子对解离为例，根据CN⁺离子出现能测定其解离阈值为13.78 eV，同时在16.09 eV获得了CN⁺离子的时间切片速度影像且显示出动量分布的各向异性.     

Physics-based plasticity model incorporating microstructure changes for severe plastic deformation

During machining processes, materials undergo severe deformations that lead to different behavior than in the case of slow deformation. The microstructure changes, as a consequence, affect the materials properties and therefore influence the functionality of the component. Developing material models capable of capturing such changes is therefore critical to better understand the interaction process–materials. In this paper, we introduce a new physics model associating Mechanical Threshold Stress (MTS) with Dislocation Density (DD) models. The modeling and the experimental results of a series of large strain experiments on polycrystalline copper (OFHC) involving sequences of shear deformation and strain rate (varying from quasi-static to dynamic) are very similar to those observed in processes such as machining. The Kocks–Mecking model, using the mechanical threshold stress as an internal state variable, correlates well with experimental results and strain rate jump experiments. This model was compared to the well-known Johnson–Cook model that showed some shortcomings in capturing the stain jump. The results show a high effect of rate sensitivity of strain hardening at large strains. Coupling the mechanical threshold stress dislocation density (MTS–DD), material models were implemented in the Abaqus/Explicit FE code. The model shows potentialities in predicting an increase in dislocation density and a reduction in cell size. It could ideally be used in the modeling of machining processes.     

On the threshold problem for Latin boxes

Let m ≤ n ≤ k. An m × n × k 0‐1 array is a Latin box if it contains exactly m n ones, and has at most one 1 in each line. As a special case, Latin boxes in which m = n = k are equivalent to Latin squares. Let be the distribution on m × n × k 0‐1 arrays where each entry is 1 with probability p, independently of the other entries. The threshold question for Latin squares asks when contains a Latin square with high probability. More generally, when does support a Latin box with high probability? Let ε > 0. We give an asymptotically tight answer to this question in the special cases where n = k and , and where n = m and . In both cases, the threshold probability is . This implies threshold results for Latin rectangles and proper edge‐colorings of K_n,n.     

Influence of temperature and alloying elements on the threshold displacement energies in concentrated Ni–Fe–Cr alloys

Concentrated solid-solution alloys(CSAs) have demonstrated promising irradiation resistance depending on their compositions. Under irradiation, various defects can be produced. One of the most important parameters characterizing the defect production and the resulting defect number is the threshold displacement energies(E_d). In this work, we report the results of E_dvalues in a series of Ni–Fe–Cr concentrated solid solution alloys through molecular dynamics(MD)simulations. Based on several different empirical potentials, we show that the differences in the E_dvalues and its angular dependence are mainly due to the stiffness of the potential in the intermediate regime. The influences of different alloying elements and temperatures on E_dvalues in different CSAs are further evaluated by calculating the defect production probabilities. Our results suggest a limited influence of alloying elements and temperature on E_dvalues in concentrated alloys. Finally, we discuss the relationship between the primary damage and E_dvalues in different alloys. Overall, this work presents a thorough study on the E_dvalues in concentrated alloys, including the influence of empirical potentials,their angular dependence, temperature dependence, and effects on primary defect production.