基于抖动测试的AFDX端到端延迟紧性修正

AFDX引入虚拟链路(Virtual Link)实现物理带宽资源的逻辑分隔。由于数据帧的异步到达和多路复用输出造成虚拟链路的时延抖动现象,并最终导致流量端到端延迟分析的不确定性。本文提出了一种基于抖动测试值的网络演算紧缩方法。通过分布式测试,获得虚拟链路在网络中的实际传输抖动,并以此为基础,建立了流量传输精确化模型,通过流量模型的逐级修正,使端到端延迟计算结果逐级精确化。通过将抖动实际值与理论分析结果相结合,提供了网络演算悲观度及其扩散影响度量的直观对比,提高了延迟计算的紧性。     

基于网络层析成像的光网络流量矩阵估计方法(英文)

提出一种面向光传输网络的流量矩阵估计方法.采用压缩感知理论研究光传输网络中的流量矩阵估计,根据信号稀疏表示将流量矩阵稀疏化,基于矩阵变换理论提出新的面向光传输网络的网络层析成像模型.该模型克服了已有网络层析成像模型的病态特性,并通过凸优化来获得流量矩阵的估计等式.提出了具体的估计算法,获得关于光传输网络流量矩阵的精确估计.真实网络的数据仿真表明所提出的方法是有效和可行的.     

自调式J-T制冷器流量测量的流量计选用

针对自调式J-T制冷器的流量测量,介绍了几种常用的气体流量计,结合制冷器的实际需求对其进行了分析,认为从进气端测量科里奥利质量流量计较为合适,而在排气端测量选用湿式气体流量计较为合适。     

提取葡萄糖多晶光学参数的迭代和遗传算法比较分析

基于太赫兹反射时域光谱技术,分别利用迭代算法和遗传算法提取葡萄糖多晶太赫兹频段的两个光学参数,即折射率和吸收系数,并将遗传算法与迭代算法进行对比分析。研究结果表明：当将利用弱吸收近似条件所获得的光学参数作为迭代算法的初始值时,采用迭代算法可以提高计算效率。然而,迭代算法对初始值敏感,当初始值偏离实际值较大时,迭代算法不能保证获得较高精度的光学参数;遗传算法通过参数编码、初始种群、遗传操纵、参数控制和条件约束等设计,保证了算法的收敛性和种群多样性。而且,遗传算法的最优解并不依赖于初始种群。与迭代算法相比,采用遗传算法所获得的物质的光学参数拥有更高的精度。因此,在基于THz光谱技术提取物质的光学参数时,本文建议采用智能优化的算法,这样可以获得高精确度的光学参数。     

基于点和直线段对应的扩展正交迭代位姿估计算法

正交迭代算法是基于点特征,全局且快速收敛的位姿估计算法,是目前性能最优的实时位姿估计算法之一.对摄像机观测到的直线段提出了其不确定性描述的新方法:目标空间直线段误差,并把该误差融入到正交迭代算法的位姿计算过程中,形成扩展正交迭代算法.该算法可以同时利用点和直线段特征.算法先把点共面性方程表达为与正交迭代算法中点共线性方程一致的数学形式,再根据目标空间点共面性误差定义了目标空间直线段共面性误差.接着把目标空间点共线性误差和直线段共面性误差二者之和作为误差函数.最后推导出使该函数最小化的迭代求解过程.实验结果表明,本文算法是有效的,精确的.与正交迭代算法的计算结果对比,由于可以同时利用直线段特征,扩展正交迭代算法的位姿估计误差更低,抗噪声性能更强,算法更稳定.     

基于贝叶斯推理的像元内部端元选择模型

提出了利用贝叶斯推理选择混合像元内端元的模型.考虑到端元光谱的不确定性,基于贝叶斯推理和线性光谱混合模型得到了像元内端元集合的后验概率表达式.在获得端元共存的先验知识基础上,结合端元光谱的正态分布函数,通过最大后验概率得到最佳的端元集合.通过对包含147431个像元的ETM+影像试验表明,相对于IDRISI软件的MRES和PG算法,该算法可削减至少70%的冗余端元,使端元选择错误导致的分解误差降低至少28%.结果表明,由于充分考虑端元光谱的小确定性和端元的共存性,通过贝叶斯推理可以大幅度提高端元选择的正确率,从而改善混合像元的分解精度.     

遗传算法用于傅里叶变换红外光谱的定量解析

本文利用遗传算法对混叠的傅里叶变换红外光谱图进行了定量的解析。文中分析了最多包含了十个组分的大气有毒有机物（苯，四氯化碳，氯苯，硝基苯，苯酚，甲苯，甲醇，邻甲酚，间甲酚和对甲酚）的混合谱图，建立了适当目标函数，使用了自适应变异技术，讨论了交叉和变异概率对结果的影响。结果证明遗传算法有着极好的非线性特征。在适当的条件下，结果的相对误差小于1％。     

基于遗传算法的飞机俯仰控制参数优化设计及重构

遗传算法不受问题性质 ( 如连续性,可微性) 的限制,能够处理传统优化算法难以解决的复杂问题,故近年来在控制参数优化方面得到了广泛的应用;但算法中,交叉概率与变异概率的选择没有给出通用标准,通过多种交叉概率与变异概率组合的优化结果比较确定合适的交叉概率与变异概率,然后从几种控制系统常用性能指标中选择最合适的一种做为适应度函数,优化得无故障时控制参数,按照重构后系统开环传递函数保持不变,即系统闭环传函极点不发生变化的原则,由升降舵损伤程度和无损伤时控制参数可得重构后的控制参数;由仿真实验可得,此方法能够较快较精确地确定控制参数,系统无超调,控制效果良好,重构后,系统性能能够保持不变,达到良好的重构效果。     

基于点线对应的多摄像机全局位姿估计迭代算法

为了解决工业视觉场景中点和直线段共存情况下的多摄像机位姿估计问题,提出了基于点线特征对应的多摄像机全局位姿估计算法.以正交迭代算法为基础,使用多个固定在平台上的已标定摄像机间的几何约束,以全局方式处理多摄像机的重叠和非重叠视图的点和直线段特征,获得更加稳定精确的位姿估计.该算法先把所有摄像机数据进行统一表达,再把摄像机观测到的全部点特征的目标空间共线性误差和直线段特征的目标空间共面性误差之和作为误差函数,最后经数学推导得到使该误差函数最小化的迭代求解过程.实验结果验证了该算法的有效性以及优越性.     

涡轮叶片端壁槽缝气与离散气膜孔冷却特性实验研究

真实发动机涡轮叶片端壁为曲面造型,并且其冷却受槽缝气、泄漏流和离散气膜冷却多种冷却气叠加影响,同时又受到主流二次流影响,因此呈现复杂冷却特性。为研究接近真实发动机涡轮叶片端壁构型和工况下的气膜冷却特性,本文采用高速风洞(主流雷诺数为37万)及压敏漆(PSP)技术,研究了槽缝气、泄漏流以及离散气膜对曲面端壁的气膜冷却效率的影响,并针对不同冷气流量比对端壁气膜冷效的影响规律进行了对比分析。结果表明:端壁表面气膜冷效随着槽缝气流量比增大而增大,当流量比增大到1.71%时,槽缝气膜几乎可以覆盖整个端壁表面;与槽缝气相比,端壁表面的离散气膜冷气覆盖范围较为有限,端壁压力面侧下游区域气膜覆盖较差;在喉部之前,随着流量比增大,离散气膜冷效呈现下降趋势;在喉部之后,随着流量比从1.3%增大到1.9%,离散气膜冷效呈现上升趋势;与仅有离散气膜相比,包含槽缝气、泄漏流、离散气膜的全气膜覆盖更为均匀,全气膜冷效的叠加使得端壁冷效相比仅有离散气膜时整体提高了93.4%。     

Driver choice compared to controlled diversion for a freeway double on-ramp in the framework of three-phase traffic theory

Two diversion schemes that apportion demand between two on-ramps to reduce congestion and improve throughput on a freeway are analyzed. In the first scheme, drivers choose to merge or to divert to a downstream on-ramp based on information about average travel times for the two routes: (1) merge and travel on the freeway or (2) divert and travel on a surface street with merging downstream. The flow, rate of merging at the ramps, and the travel times oscillate strongly, but irregularly, due to delayed feedback. In the second scheme, diversion is controlled by the average mainline velocities just upstream of the on-ramps. Driver choice is not involved. If the average upstream velocity on the mainline drops below a predetermined value (20 m/s) vehicles are diverted to the downstream ramp. When the average mainline velocity downstream becomes too low, diversion is no longer permitted. The resultant oscillations in this scheme are nearly periodic. The period is dominated by the response time of the mainline to interruption of merging rather than delayed feedback, which contributes only a minor component linear in the distance separating the on-ramps. In general the second scheme produces more effective congestion reduction and greater throughput. Also the travel times for on-ramp drivers are less than that obtained by drivers who attempt to minimize their own travel times (first scheme). The simulations are done using the Kerner-Klenov stochastic three-phase theory of traffic [B.S. Kerner, S.L. Klenov, Phys. Rev. E 68 (2003) 036130].     

Traffic noise reduction due to the porous road surface

Porous road surfaces reduce road traffic noise. A new method of noise reduction assessment is proposed. The noise generated by a few vehicles was measured two times: on an old surface with the dense asphalt and on a new surface with the porous asphalt. Subjective assessments of drive-by noise suggest that the sound exposure and the road surface coefficient can be used as the acoustical characteristics of a road surface. Their average values, with the average number of vehicles passing the receiver during a day or night, makes it possible to predict the equivalent continuous A-weighted sound pressure level for the new road surface. This is the main objective of this paper.     

Effect of adaptive cruise control systems on mixed traffic flow near an on-ramp

Mixed traffic flow consisting of vehicles equipped with adaptive cruise control (ACC) and manually driven vehicles is analyzed using car-following simulations. Simulations of merging from an on-ramp onto a freeway reported in the literature have not thus far demonstrated a substantial positive impact of ACC. In this paper cooperative merging for ACC vehicles is proposed to improve throughput and increase distance traveled in a fixed time. In such a system an ACC vehicle senses not only the preceding vehicle in the same lane but also the vehicle immediately in front in the other lane. Prior to reaching the merge region, the ACC vehicle adjusts its velocity to ensure that a safe gap for merging is obtained. If on-ramp demand is moderate, cooperative merging produces significant improvement in throughput (20%) and increases up to 3.6 km in distance traveled in 600 s for 50% ACC mixed flow relative to the flow of all-manual vehicles. For large demand, it is shown that autonomous merging with cooperation in the flow of all ACC vehicles leads to throughput limited only by the downstream capacity, which is determined by speed limit and headway time.     

Noise variability due to traffic spatial distribution

The inverse relation of the uncertainty of the equivalent sound pressure level to the square root of the number of vehicle pass-bys is investigated. The influence of the traffic spatial distribution is evaluated by means of a computer model, based on Monte Carlo and the Weyl–Ingard Theory. The model results are discussed and compared with experimental data. By least mean squares fitting of simulated results, a new expression is obtained for the sound pressure level uncertainty, that takes into account the receiver position, road configuration and traffic flow balance. This expression explains the different values found in the literature and yields uncertainty values that may not be covered by the ISO 1996-2:2007 recommendation, in some specific circumstances. Finally the path for building traffic noise uncertainty maps is sketched out.     

Application of simulation program to specific urban situation

In the paper, the simulation program predicting the sound level has been applied to one of buildings neighbouring a road. In the simulation are considered: geometry of road surroundings, the shape of the investigated building façade including, road geometry (number of lanes and their positions) and traffic structure (vehicle flow rates and their average speeds). Adjustable accuracy in an urban system modelling allows finding the model which is adequate to the real system. For the model of the accuracy limited by the simulation error due to vehicle speed estimation, by comparison with the field measurement, the model validation range has been established. Based on the simulation adjustable accuracy, possibilities of the validation range extension have been analysed.     

Universal behavior in a model of city traffic with unequal green/red times

The effect of green/red asymmetry is studied for the single-car traffic model proposed in [B.A. Toledo, V. Muñoz, J. Rogan, C. Tenreiro, J.A. Valdivia, Modeling traffic through a sequence of traffic lights, Phys. Rev. E 70 (1) (2004) 016107], on two different signal synchronization strategies, namely, all signals in phase, and a green wave. The asymmetry is characterized by the parameter g=_tgr/T$g=t_{\text{gr}}/T$, where _tgr$t_{\text{gr}}$ is the green time and T$T$ the signal period. Although the car dynamics turns simpler or more complex, as compared with the equivalent situation for the symmetric case g=0.5$g=0.5$, critical behavior around resonance is shown to be preserved. However, unlike the case g=0.5$g=0.5$, critical exponents at both sides of the resonance are not equal and depend on g$g$. Analytical expressions for them are found, and shown to be both consistent with simulation results and independent of the distribution of distances between signals for the green wave case. Also, it is found that the green wave strategy is more robust to changes in g$g$, with respect to the synchronized lights strategy, in the sense that larger departures from g=0.5$g=0.5$ are needed to have noticeable effects on the car dynamics.     

Multiple jamming transitions in traffic flow

The effect of accelerating stepwise on the jamming transition is investigated in the extended car-following model. The optimal velocity function is modified to take into account accelerating stepwise vehicles. It is shown that the multiple phase transitions occur on varying the car density. The multiple transitions change with the delay time. The flow-density curves and the velocity-headway curves are presented for various delay times. It is also shown that the multiple jamming transition lines are consistent with the neutral stability curves. The jamming transitions are closely related with the turning points of the optimal velocity function.     

Estimation of the long term average sound level from hourly average sound levels

It is shown how to estimate the long-term average sound level, L_AeqLT (for free flowing road traffic) from measurements of the hourly A-weighted equivalent sound level, L_Aeq1h. To estimate the parameters of the model which describe noise emission and attenuation, concurrent measurements of L_Aeq1h at two distances from the considered road are needed. A semi-empirical formula is derived for L_AeqLT approximation. Also the uncertainty of this approximation is given as a function the distance from the road and receiver height.     

A new lattice model of traffic flow with the consideration of the traffic interruption probability

In this paper, we present a new lattice model which involves the effects of traffic interruption probability to describe the traffic flow on single lane freeways. The stability condition of the new model is obtained by the linear stability analysis and the modified Korteweg-de Vries (KdV) equation is derived through nonlinear analysis. Thus, the space will be divided into three regions: stable, metastable and unstable. The simulation results also show that the traffic interruption probability could stabilize traffic flow.     

A quantum mechanics-based approach to model incident-induced dynamic driver behavior

A better understanding of the psychological factors influencing drivers, and the resulting driving behavior responding to incident-induced lane traffic phenomena while passing by an incident site is vital to the improvement of road safety. This paper presents a microscopic driver behavior model to explain the dynamics of the instantaneous driver decision process under lane-blocking incidents on adjacent lanes. The proposed conceptual framework decomposes the corresponding driver decision process into three sequential phases: (1) initial stimulus, (2) glancing-around car-following, and (3) incident-induced driving behavior. The theorem of quantum mechanics in optical flows is applied in the first phase to explain the motion-related perceptual phenomena while vehicles approach the incident site in adjacent lanes, followed by the incorporation of the effect of quantum optical flows in modeling the induced glancing-around car-following behavior in the second phase. Then, an incident-induced driving behavior model is formulated to reproduce the dynamics of driver behavior conducted in the process of passing by an incident site in the adjacent lanes. Numerical results of model tests using video-based incident data indicate the validity of the proposed traffic behavior model in analyzing the incident-induced lane traffic phenomena. It is also expected that such a proposed quantum-mechanics based methodology can throw more light if applied to driver psychology and response in anomalous traffic environments in order to improve road safety.