对四类泻流问题的深入研究

对热学中的4类泻流问题进行了详细的计算,并对结果进行了分析讨论.恒温的泻流问题有解析的解,而绝热泻流问题往往受限于数学知识只能得到数值的结果;同时本文也对泻流的近平衡态条件进行了大致的分析.这些将有助于广大师生对于泻流问题的全面认识.     

Liner ship fleet deployment with week-dependent container shipment demand

This paper addresses a practical liner ship fleet deployment problem with week-dependent container shipment demand and transit time constraint, namely, maximum allowable transit time in container routing between a pair of ports. It first uses the space–time network approach to generate practical container routes subject to the transit time constraints. This paper proceeds to formulate the fleet deployment problem based on the practical container routes generated. In view of the intractability of the formulation, two relaxation models providing lower bounds are built: one requires known container shipment demand at the fleet deployment stage, and the other assumes constant container shipment demand over the planning horizon. An efficient global optimization algorithm is subsequently proposed. Extensive numerical experiments on the shipping data of a global liner shipping company demonstrate the applicability of the proposed model and algorithm.     

集装箱翻箱问题的蚁群算法改进

翻箱问题属于NP难问题,基本蚁群算法在求解该问题上收敛困难且寻优能力低。因此,本文提出了一种适合于翻箱模型的改进型蚁群算法,在概率决策机制、解的重构、信息素更新机制三个方面对基本蚁群算法进行改进。最后通过与其他算法的分析比较,验证了该改进算法的可行性与有效性。     

储液容器跌落冲击动力学神经网络建模

运用LS-DYNA程序中的ALE算法模拟储液容器在不同的跌落角度、跌落高度、壳体厚度下的跌落冲击过程,获取神经网络预测模型的训练样本集;利用BP神经网络建立储液容器结构参数、跌落冲击参数与接触点最大应力之间的映射关系预测模型,并将各种参数下的接触点最大应力网络预测值与仿真值比较,两者差异较小,表明该方法是有效的,可以为实际生产过程中参数选择提供理论依据.     

用于ALE有限元模拟的网格更新方法

任意拉格朗日欧拉法(ALE)可以通过定义参考网格的运动,实现自由液面跟踪,完成液体晃动的数值计算. 综合用于更新网格节点的3种基本计算方法,将多方向更新网格速度的技术应用于任意拉格朗日欧拉网格节点的速度计算. 给出了水平圆柱形贮箱和椭圆形贮箱内液体晃动算例,实现了多方向更新网格运动与晃动流场计算的耦合,使ALE方法能胜任复杂几何边界下的自由液面流动的数值模拟.      

爆炸容器内冲击波系演化及壳体响应的数值研究

对中心装药爆炸后冲击波的产生、传播和壳体动态响应全过程进行了数值研究。认为RDX瞬时爆炸 ,爆炸近场采用自相似解 ;冲击波传播和波系演化采用PPM (the Piecewise Parabolic Method)格式求解Eu ler方程 ;壳体响应采用有限元方法求解拉氏坐标系下由虚功原理得到的动力学方程。壳体内壁面边界条件分别采用强耦合和弱耦合方法处理。结果表明 :(1)当装药量相同时 ,薄壁壳体振型比厚壁壳体复杂得多 ,振幅也大 ;(2 )当装药量不同 ,壳体厚度相同时 ,爆炸场冲击波的演化过程不同 ;(3)对少量装药 ,产生的冲击波强度低 ,壳体变形小 ,是否考虑内边界运动 ,对计算结果的影响不大 ;(4 )在本文条件下 ,爆炸容器封头顶点所受的载荷最大 ,是最易发生破坏的地方 ,侧壁与爆点所在横截面的交线 ,也易破坏。     

旋转液体中固体颗粒运动的进一步讨论

通过分析旋转液体中的压强分布,说明了浸在液体中的固体颗粒向轴心运动的原因,以及该规律的实际应用.     

A Tetrairon Dication Featuring Tetraethynylbenzene Bridging Ligand: A Molecular Prototype of Quantum Dot Cellular Automata

The unprecedented tetrairon dication [{Cp*(dppe)FeC≡C-}₄-μ-(1,2,4,5-C₆H₂)](PF₆)₂ ( 1 ) was obtained through a sequence of three reactions from 1,2,4,5-tetraethynylbenzene, Cp*(dppe)FeCl (Cp*=C₅Me₅, dppe=1,2-bis(diphenylphosphino)-ethane), KOtBu, and ferrocenium hexafluorophosphate. The cyclic voltammogram of the target molecule, isolated in 77 % yield, exhibits four well separated and reversible redox events showing that 1 is thermodynamically stable with respect to disproportionation (Kc>10⁶). The tetranuclear dication 1 was characterized by XRD on single crystal, IR and NMR spectroscopies and Mössbauer spectrometry. The experimental data show that 1 behaves as a class II mixed-valence complex with the positive charges preferentially disposed on antipodal positions. This new molecule can be regarded as a potential molecular prototype of quantum dot cellular automata.     

The service allocation problem at the Gioia Tauro Maritime Terminal

The Service Allocation Problem (SAP) is a tactical problem arising in the yard management of a container transshipment terminal. The objective is the minimization of the container rehandling operations inside the yard. This study of the SAP was undertaken for the Gioia Tauro port which is located in Italy and is the main hub terminal for container traffic in the Mediterranean Sea. The SAP can be formulated as a Generalized Quadratic Assignment Problem (GQAP) with side constraints. Two mixed integer linear programming formulations are presented. The first one exploits characteristics of the yard layout at Gioia Tauro where the berth and the corresponding yard positions extend along a line. The second formulation is an adaptation of a linearization for the GQAP. In both cases only small instances can be solved optimally. An evolutionary heuristic was therefore developed. For small size instances the heuristic always yields optimal solutions. For larger sizes it is always better than a truncated branch-and-bound algorithm applied to the exact formulations.