基于径向基函数网络的水声信号预测模型研究

随着舰船隐身降噪技术的不断发展,舰船辐射噪声信号检测变得愈加困难.针对这种情况,将RBF神经网络引入到水声信号的预测中,建立水声信号的全局预测模型.通过对Henon和Lorenz映射仿真,验证了RBF预测模型的有效性.运用全局预测模型对两种实际水声信号进行预测.实验证明,该预测模型学习速度快,所需样本点少,效果较准确,在水声信号检测中具有良好的发展前景.     

Evaluation of consequence assessment methods for pool fires on water involving large spills from liquefied natural gas carriers

This article concerns thermal radiation hazards associated with unconfined liquefied natural gas (LNG) spills on water. Consequence assessment methods were compared to clarify their model characteristics in large-scale LNG spills from an LNG carrier (LNGC). The consequences of LNG release, pool spread, and pool fire hazards were estimated using the following practical methods: the Federal Energy Regulatory Commission’s (FERC) method, the Sandia National Laboratories’ method, and the Fay method. The sensitivity of consequence analysis results to the breach size of a tank was examined under the assumption that LNG is released from a common type LNGC of 125000 m³ cargo capacity. Consequently, it was found that the FERC method is useful from the practical viewpoint of being applicable to any breach size. Finally, thermal radiation hazards from pool fires involving spills from one of the latest and largest LNGCs (250 000 m³ cargo capacity), which are currently considered for construction, were investigated using the recommended FERC method, and the results are discussed in comparison with those for common type LNGCs. As a result, it was found that the maximum thermal hazard distance is longer by only about 24% compared with the common type LNGC, whereas the spill volume is twice as much.     

高原公路水土流失预测

以香日德-花石峡高原项目区土壤侵蚀现状为例,结合3种土壤侵蚀的特点,按照水土保持技术规范,提出分区、分类的高原公路水土流失预测方法,并详细分析预测结果,提出相应的监测和治理建议.     

车门密封胶条力学性能衰变分析

按照国家标准的规定,检测自然条件使用的密封胶条样品的硬度和拉伸性能(拉伸强度、拉断伸长率),从自然老化的角度分析了密封胶条老化前后力学性能变化规律,从性能保持率的角度分析了密封胶条的劣化程度,最后基于密封胶条拉断伸长率的数据,构建了6阶拉格朗日插值多项式,对车门密封胶条使用寿命进行了预测,结果显示其使用寿命是185个月。     

Time domain modeling and simulation of warship flooding process based on counter-flooding activities北大核心CSCD

[目的]实战损伤条件下,舰艇进水是舱室进水漫延与舰员抗沉行为相对抗的过程。现有的不沉性理论无法客观、全面评估抗沉干预对战损舰艇进水过程的影响,缺少从全时域的角度对战损舰艇进水过程开展预测的手段。[方法]首先,分析抗沉干预下舰艇破损进水的耦合关系,构建进水过程时域模型,进而给出数值求解算法;其次,结合案例分析抗沉干预要素对不沉性的影响,并开发可视化、交互式的仿真系统。[结果]通过仿真系统实现了进水过程与多种抗沉干预方式的综合评估,是对现有不沉性理论的有效补充。[结论]该成果不仅可帮助设计人员掌握抗沉设计方案对干预进水过程的有效性,指导舰艇抗沉能力的设计;还可应用于抗沉指挥决策的态势预报,具有较高的军事应用价值。     

断裂力学在轨道车辆结构寿命预测中的应用

轨道车辆结构在服役过程中难免会产生缺陷。为确保含缺陷车辆结构的运行安全,文章引入针对缺陷的断裂力学疲劳分析方法和结构完整性评估标准BS 7910,介绍了其特点,并阐述了基于BS 7910标准对含缺陷车辆结构进行寿命预测的关键环节。根据预测结果,可指导运维单位对结构制定维保策略。     

动荷载作用下钢结构涂装的累积损伤机理与量化模型

为研究动荷载作用下公路钢桥涂装层与基体间的协同变形和破坏机制,开发了钢结构涂装层试件的动荷载试验系统,研究了疲劳荷载作用下涂层的破坏模式、附着力、弹性厚度、应变等随循环加载次数的发展规律。结合涂层破坏模式和发展规律,构建了涂层疲劳累积损伤度的计算模型,提出了基于损伤度的涂层破坏分级方法,并结合试验结果进行了验证。结果表明:疲劳荷载将引起涂层的附着力下降、弹性厚度降低和应变松弛,进而导致涂层出现附着破坏、内聚破坏及其组合。涂层附着损伤度是涂层损伤的主要因素,且随加载次数持续增大,当加载次数较大时内聚损伤贡献突出。试验涂装体系在22万次应力循环后附着损伤度最大为0.34,内聚损伤度最大为0.27,总损伤度达0.61,属于严重损伤。提出的损伤度模型及破坏分级方法与试验规律吻合良好,可应用于公路钢桥涂装层受动荷载作用下的寿命预测。     

物流无人机对地风险评估方法研究

随着智能交通运输体系的完善和无人机应用场景的拓展，物流无人机应运而生。物流无人机被认为是解决快递配送“最后一公里”问题的有效运输方式，但由空中碰撞、系统失效等原因导致的无人机坠落有可能对地面人员造成伤害并产生经济损失。因此本文聚焦无人机物流运输场景，研究物流无人机坠地后造成的死亡人数和经济损失，并利用风险矩阵综合评估风险。主要内容包括：分析物流无人机空中碰撞后产生机货分离的情况，建立由物流无人机和货物坠地造成 的地面死亡人数计算模型；对无人机和货物进行分类，建立经济损失计算模型，包括由无人机和货物价值损失构成的直接经济损失，以及由社会服务机构费用消耗构成的间接经济损失；综合物 流无人机和货物坠地造成的地面死亡人数和经济损失建立风险评估矩阵。算例证明，利用本文方法得到的地面死亡人数为每飞行小时3.74×10-12~1.87×10-7人，符合等效安全水平；每次事故经济损失为6734.8~33619.0元，也在合理区间内。本文提供的方法可以对物流无人机地面风险评估和管控提供一定借鉴和指导意义。     

武器装备的可靠性再分配方法及应用探讨

文章讨论了如何充分利用可靠性预计值来优化可靠性分配的问题。在已知可靠性预计值和初始分配值时,视情况对可靠性进行再分配。并用实例说明此算法的应用和合理性。     

The effect of railway local irregularities on ground vibration

The environmental effects of ground-borne vibrations generated due to localised railway defects is a growing concern in urban areas. Frequency domain modelling approaches are well suited for predicting vibration levels on standard railway lines due to track periodicity. However, when considering individual, non-periodic, localised defects (e.g. a rail joint), frequency domain modelling becomes challenging. Therefore in this study, a previously validated, time domain, three-dimensional ground vibration prediction model is modified to analyse such defects. A range of different local (discontinuous) rail and wheel irregularity are mathematically modelled, including: rail joints, switches, crossings and wheel flats. Each is investigated using a sensitivity analysis, where defect size and vehicle speed is varied. To quantify the effect on railroad ground-borne vibration levels, a variety of exposure–response relationships are analysed, including: peak particle velocity, maximum weighted time-averaged velocity and weighted decibel velocity. It is shown that local irregularities cause a significant increase in vibration in comparison to a smooth track, and that the vibrations can propagate to greater distances from the line. Furthermore, the results show that step-down joints generate the highest levels of vibration, whereas wheel flats generate much lower levels. It is also found that defect size influences vibration levels, and larger defects cause greater vibration. Lastly, it is shown that for different defect types, train speed effects are complex, and may cause either an increase or decrease in vibration levels.