基于QoS的卫星网络端-端通信可靠性分析

针对卫星网络动态环境下的高速信息传输、业务类型差异大等特点，提出一种综合考虑各业务QoS（Quality of Service）指标的可靠性分析方法。在卫星通信网络实际运行周期内，通信系统往往处于逐渐劣化过程中，导致卫星的节点和链路除正常工作和完全失效外，还存在部分失效的工作状态。本文在链路多状态基础上基于最小路集算法（Minimum Path Set Algorithms，MPSA）在不同业务的QoS指标（时延、带宽和丢包率）约束下，得出满足该业务QoS约束的所有可靠路径集，对路径集中路径进行不交化处理得到网络端-端可靠性。研究结果表明，不同业务由于QoS需求的差异导致网络端-端可靠性不同，所提算法与传统算法相比更加符合实际。由于实际卫星网络环境中会采用端-端并行多路径传输（Multi-Path Transmission，MTP），本文在上述研究的基础上，进一步对多路径的端-端可靠性进行了研究，结果表明多路径数据传输可靠性高。     

Advanced multiple response surface method of sensitivity analysis for turbine blisk reliability with multi-physics coupling

To reasonably implement the reliability analysis and describe the significance of influencing parameters for the multi-failure modes of turbine blisk, advanced multiple response surface method(AMRSM) was proposed for multi-failure mode sensitivity analysis for reliability. The mathematical model of AMRSM was established and the basic principle of multi-failure mode sensitivity analysis for reliability with AMRSM was given. The important parameters of turbine blisk failures are obtained by the multi-failure mode sensitivity analysis of turbine blisk. Through the reliability sensitivity analyses of multiple failure modes(deformation, stress and strain) with the proposed method considering fluid–thermal–solid interaction, it is shown that the comprehensive reliability of turbine blisk is 0.9931 when the allowable deformation, stress and strain are3.7*10~(-3)m, 1.0023*10~9 Pa and 1.05*10~(-2)m/m, respectively; the main impact factors of turbine blisk failure are gas velocity, gas temperature and rotational speed. As demonstrated in the comparison of methods(Monte Carlo(MC) method, traditional response surface method(RSM), multiple response surface method(MRSM) and AMRSM), the proposed AMRSM improves computational efficiency with acceptable computational accuracy. The efforts of this study provide the AMRSM with high precision and efficiency for multi-failure mode reliability analysis, and offer a useful insight for the reliability optimization design of multi-failure mode structure.     

Dynamics of air transport networks: A review from a complex systems perspective

Air transport systems are highly dynamic at temporal scales from minutes to years. This dynamic behavior not only characterizes the evolution of the system but also affect the system's functioning. Understanding the evolutionary mechanisms is thus fundamental in order to better design optimal air transport networks that benefits companies, passengers and the environment. In this review, we briefly present and discuss the state-of-the-art on time-evolving air transport net-works. We distinguish the structural analysis of sequences of network snapshots, ideal for long-term network evolution (e.g. annual evolution), and temporal paths, preferred for short-term dynamics (e.g. hourly evolution). We emphasize that most previous research focused on the first modeling approach (i.e. long-term) whereas only a few studies look at high-resolution temporal paths. We conclude the review highlighting that much research remains to be done, both to apply already available methods and to develop new measures for temporal paths on air transport networks. In particular, we identify that the study of delays, network resilience and optimization of resources (aircraft and crew) are critical topics.     

基于蒙特卡洛仿真的平行双跑道容量评估方法

跑道容量评估是科学评价跑道规划方案、合理安排航班时刻、精准实施流量管理的重要基础性工作.针对平行双跑道容量评估问题,分析了跑道容量影响因素,考虑到跑道分配、机型分配、起降类型分配、进离港点分配、航班流队列顺序等随机性,采用蒙特卡洛方法模拟变量随机性,按照离散事件系统原理,递归计算每个航班的起降时刻,通过统计1h内的航班数量获得容量值.以西安机场双跑道为例,针对现行的相关进近、独立进近、隔离运行3种模式,分别计算了跑道容量值.仿真计算结果显示,相关进近模式容量60架次/h,独立进近模式跑道容量为78架次/h,隔离运行模式跑道容量为52架次/h,平均每次仿真时间仅为1 min,说明方法效率较高.与2015年高峰61架次/h相比仅相差1个架次,计算结果说明方法是有效的.     

高可扩展格子Boltzmann方法

格子Boltzmann方法(Lattice Boltzmann Method,LBM)是计算流体力学中的一种常用方法.由于LBM中的格点仅与相邻的格点间存在数据传递,因此具有良好的并行性.LBM并行算法中的数值通信部分通常采用的是格点上的微观量——分布函数.每次传递的分布函数具有多个不同的速度方向,为了进一步减少LBM并行算法的通信开销,从格子 Boltzmann方法的物理特性以及相应的串行程序为切入点,深层次地挖掘可并行的因子,设计了专门用于通信面的类,使用数量较少的宏观量进行通信,降低了通信所占的比重,缩短了通信时间,提高了加速比和效率.实验表明,在4 096个计算核上依然有良好的加速比和效率.     

Performance analysis of PPP ambiguity resolution with UPD products estimated from different scales of reference station networks

Integer ambiguity fixing with uncalibrated phase delay (UPD) products can significantly shorten the initialization time and improve the accuracy of precise point positioning (PPP). Since the tracking arcs of satellites and the behavior of atmospheric biases can be very different for the reference networks with different scales, the qualities of corresponding UPD products may be also various. The purpose of this paper is to comparatively investigate the influence of different scales of reference station networks on UPD estimation and user ambiguity resolution. Three reference station networks with global, wide-area and local scales are used to compute the UPD products and analyze their impact on the PPP-AR. The time-to-first-fix, the unfix rate and the incorrect fix rate of PPP-AR are analyzed. Moreover, in order to further shorten the convergence time for obtaining precise positioning, a modified partial ambiguity resolution (PAR) and corresponding validation strategy are presented. In this PAR method, the ambiguity subset is determined by removing the ambiguity one by one in the order of ascending elevations. Besides, for static positioning mode, a coordinate validation strategy is employed to enhance the reliability of the fixed coordinate. The experiment results show that UPD products computed by smaller station network are more accurate and lead to a better coordinate solution; the PAR method used in this paper can shorten the convergence time and the coordinate validation strategy can improve the availability of high precision positioning.     

Research on vibration suppression of a mistuned blisk by a piezoelectric network

The work aims to provide a further investigation of the dynamic characteristics of an integral bladed disk(also called ‘blisk') with a Parallel Piezoelectric Network(PPN). The PPN is constructed by parallelly interconnecting the piezoelectric patches distributed in the blisk. Two kinds of PPN are considered, namely mono-periodic PPN and bi-periodic PPN. The former has a piezoelectric patch in each sector, and the later has one patch every few sectors. The vibration suppression performance of both kinds of PPN has been studied through modal analysis, forced response analysis, and statistical analysis. The research results turn out that the PPN will only affect mechanical frequencies near the electrical frequency clusters slightly, and the bi-periodic PPN will make the nodal diameter spectrum of the modes more complex, but the amplitude corresponding to the new nodal diameter component is much smaller than that of the nodal diameter component corresponding to the mono-periodic system. The mechanical coupling between the blades and the disk plays an important role in the damping effect of the PPN, and it should be paid attention to in applications. The mono-periodic PPN can effectively suppress the amplitude magnification of the forced response induced by the mistuning of the blisk; meanwhile, it can mitigate the vibration localization of the mistuned electromechanical system. If piezoelectric patches are set only in part of the sectors, the bi-periodic PPN still has a vibration suppression ability, but the effect is related to the number and spatial distribution of the piezoelectric patches.     

专用指令集在基于FPGA的神经网络加速器中的应用#br#

近年来，表现出极其优越性能的神经网络算法对硬件算力的要求逐渐提高.在一些低功耗场景如星载系统中，拥有可编程重构、高并行等特性的FPGA是神经网络算法较为合适的硬件加速平台.为了解决传统神经网络硬件加速器设计中片内资源消耗大、各功能模块耦合性高等问题，设计实现了一套专用AI指令集并应用在了基于FPGA的神经网络加速器的设计中.文章首先介绍了该指令集的设计方案.整个指令集由指令寄存器、指令解释器、指令转发模块、内存管理单元和多个模块构成.通过该指令集可实现对不同模块的复用，降低模块之间的耦合性.并以YOLOV3 Tiny网络模型为例，对比了平铺式和指令控制式两种加速方案的逻辑资源的消耗.验证了应用专用指令集可以减少约50%的FPGA逻辑资源的使用.     

具有时变权矩阵的离散Hopfield神经网络的稳定性

研究了以并行工作模式进行演化的具有时变连接权矩阵的离散Hopfield神经网络,得到了判定网络稳定的2个充分条件,改进了已有结果。     

弹道跟踪数据融合处理的快速算法

针对弹道跟踪数据融合处理中的大计算量环节研究了快速算法。用样条函数表示弹道参数,建立了多测元的联合观测模型和弹道参数的非线性融合计算模型,给出了弹道参数的求解算法,分析了弹道参数融合计算中的大型矩阵运算问题,利用基础线性代数函数库提高了大型矩阵的运算速度。建立了样条模型计算的非线性约束优化模型,给出了确定样条节点位置的优化算法,通过分析样条模型的计算原理设计了并行算法,实现了样条模型的并行化计算。仿真结果表明,弹道参数融合计算和样条模型计算的效率都得到了显著提高,计算时间减少了65.47%,对缩短数据处理周期有重要意义。