Anticipative control of switched queueing systems

The relevant dynamics of a queueing process can be anticipated by taking future arrivals into account. If the transport from one queue to another is associated with transportation delays, as it is typical for traffic or productions networks, future arrivals to a queue are known over some time horizon and, thus, can be used for an anticipative control of the corresponding flows. A queue is controlled by switching its outflow between “on” and “off” similar to green and red traffic lights, where switching to “on” requires a non-zero setup time. Due to the presence of both continuous and discrete state variables, the queueing process is described as a hybrid dynamical system. From this formulation, we derive one observable of fundamental importance: the green time required to clear the queue. This quantity allows to detect switching time points for serving platoons without delay, i.e., in a “green wave” manner. Moreover, we quantify the cost of delaying the start of a service period or its termination in terms of additional waiting time. Our findings may serve as a basis for strategic control decisions.     

The Effect of Queueing Strategy on Network Traffic

In recent years, the transportation system has been faced by increasing challenge in congestion and inefficiency, and research in traffic network has become a significant area of interest. In this paper, we introduce a dynamic-information-based (DIB) queueing strategy into network traffic model under the efficient routing strategy. DIB makes a packet with higher priority to be delivered if there are less packets travelling along its path from the current node to the destination. It is found that, compared with the traditional first-in-first-out (FIFO) queueing strategy, DIB can effectively balance the traffic load of the system via delaying packets to be delivered to congested nodes. Although the network capacity has no obvious changes, some other indexes which reflect transportation efficiency are efficiently improved in the congestion state. Besides, extensive simulation results and discussions are provided to explain the phenomena. The results may provide novel insights for research on traffic systems.     

Congestion analysis of traffic networks with direction-dependant heterogeneity

Traffic flow directionality and network weight asymmetry are widespread notions in traffic networks. This paper investigates the influence of direction-dependant heterogeneity on traffic congestion. To capture the effect of the link directionality and link weight asymmetry, the heterogeneity indexes of complex networks and the traffic flow model are introduced. The numerical results show that the critical value of heterogeneity determines congestion transition processes. The congestion degree increases with heterogeneity when the network heterogeneity is at a subcritical region. A network is more tolerant of congestion if the heterogeneity of the network is smaller or larger than the critical value. Furthermore, when heterogeneity reaches the critical value, the average number of accumulated vehicles arrives at the maximum and the traffic flow is under a serious congestion state. A significant improvement on the tolerance to congestion of traffic networks can be made if the network heterogeneity is controlled within a reasonable range.     

Magnetic bistability and controllable reversal of asymmetric ferromagnetic nanorings

Magnetization reversals through the formation of a vortex state and the rotation of an onion state are two processes with comparable probabilities for symmetric magnetic nanorings with a radius of about 50 nanometers. This magnetic bistability is the manifestation of the competition between the exchange energy and the magnetostatic energy in nanomagnets. The relative probability of the two processes in symmetric nanorings is dictated by the ring geometry and cannot be altered after fabrication. In this work, we report a novel type of nanorings--asymmetric nanorings. By tuning the asymmetry, we can control the fraction of the vortex formation process from about 40% to nearly 100% by utilizing the direction of the external magnetic field. The observed results have been accounted for by the dependence of the domain-wall energy on the local cross-section area for which we have provided theoretical calculations.     

压缩真空中的三能级原子的自发辐射

对压缩真空与Ｖ型三能级原子的相互作用作了理论分析,探讨了量子干涉效应对压缩真空中原子自发辐射的影响。当原子的上能级简并时强的量子干涉效应影响原子的定态和定态出射光谱,使它们对原子的初态敏感。这种情况下,压缩使谱线趋于关于压缩真空中心频率对称,量子干涉使谱线趋于不对称。     

Single Transverse Spin Asymmetries at Parton Level

Two factorization approaches have been proposed for single transverse spin asymmetries. One is the cofiinear factorization, the other is the transverse-momentum-dependent factorization. They have been previously derived in a formal way by using diagram expansion at hadron level. If the two factorizations hold or can be proven, they should also hold when we replace hadrons with patton states. We examine these two factorizations at patton level with massless partons. It is nontrivial to generate these asymmetries at parton level with massless patrons because the asymmetries require helicity-flip and nonzero absorptive parts in scattering amplitudes. By constructing suitable patton states with massless partons we derive the two factorizations for the asymmetry in Drell-Yan processes. It is found from our results that the collinear factorization derived at parton level is not the same as that derived at hadron level. Our results with massless partons confirm those derived with single massive parton state in our previous works.     

人脑默认模式网络的动力学行为

大脑具有自适应、自组织、多稳态等重要特征,是典型的复杂系统.人脑在静息态下的关键功能子网络--默认模式网络(DMN)的激活处于多状态间持续跳转的非平衡过程,揭示该过程背后的动力学机制具有重要的科学意义和临床应用前景.本文基于功能磁共振获得的血氧水平依赖(BOLD)信号,建立了DMN吸引子跳转非平衡过程的能量图景、吸引子非联通图、跳转关系网络等;以高级视觉皮层和听觉等皮层活动为例,通过对应激活DMN状态空间的分布,以及XGBoost、深度神经网络等算法验证了DMN状态变化与外部脑区状态的密切依赖关系;通过偏相关、收敛交叉映射等方法分析了DMN内各个脑区之间的相互作用.本文结果有助于理解静息态下大脑内在非平衡过程的动力学机制,以及从动力学的角度探索具有临床意义的脑功能障碍生物标志物.     

Matrix Representation of the Stationary Measure for the Multispecies TASEP

In this work we construct the stationary measure of the N species totally asymmetric simple exclusion process in a matrix product formulation. We make the connection between the matrix product formulation and the queueing theory picture of Ferrari and Martin. In particular, in the standard representation, the matrices act on the space of queue lengths. For N>2 the matrices in fact become tensor products of elements of quadratic algebras. This enables us to give a purely algebraic proof of the stationary measure which we present for N=3.     

Unitary state preparation,local position measurements,and spin in quantum mechanics

The orthodox presentation of quantum theory often includes statements on state preparation and measurements without mentioning how these processes can be achieved. The often quoted projection postulate is regarded by many as problematical. This paper presents a systematic framework for state preparation and measurement. Within the existing Hilbert space formulation of quantum mechanics for spinless particles we show that it is possible (1)to prepare an arbitrary state and (2)to reduce all quantum measurements to local position measurements in an asymptotic way by unitary evolution processes without recourse to the projection postulate. A generalization to spin-1/2particles is also given. The theory presented provides a general mathematical and theoretical foundation for many practical schemes for state preparation and measurement.     

Spontaneous symmetry breaking origin for the difference between time and space

We suggest that the difference between time and space is due to spontaneous symmetry breaking. In a theory with spinors the signature of the metric is related to the signature of the Lorentz group. We discuss a higher symmetry that contains pseudo-orthogonal groups with an arbitrary signature as subgroups. The fundamental asymmetry between time and space can then result as a property of the ground state rather than being put into the formulation of the theory a priori. We show how the complex structure of quantum field theory as well as gravitational field equations arise from spinor gravity--a fundamental spinor theory without a metric.     

InAs/InP柱型量子线中隧穿时间和逃逸问题的研究

在有效质量近似和绝热近似下,利用转移矩阵法研究了电子通过In As/In P/In As/In P/In As柱形量子线共振隧穿二极管的输运问题,分析和讨论了电子居留时间以及电子的逃逸过程.详细研究了外加电场、结构尺寸效应对居留时间和电子逃逸的影响.居留时间随电子纵向能量的演化呈现出共振现象;同时,结构的非对称性对电子居留时间有很大的影响,随着结构非对称性的增加,居留时间表现出不同的变化.利用有限差分方法研究了非对称耦合量子盘中电子的相干隧穿逃逸过程.     

Nondipole parameters of TDCS for electron impact ionization and drag current

A comprehensive study is undertaken of angular distributions of electron knock-out from atomic targets by fast electrons with a small transfer of momentum. The general expressions for the parameters of the triple differential cross-section of impact ionization in the optical limit are derived. The calculated parameters are compared with those of the angular distribution of electrons ejected from an atom in the process of photoionization. In these processes, when the multipole transitions are involved, the one-to-one correspondence between the photoionization and impact ionization parameters disappears. The nondipole transitions lead to the backward/forward asymmetry of the angular distribution of ejected electrons that is absent in the dipole approximation for ionization by both fast electrons and photons. Using the He atom as an example, the character of the asymmetry for these two processes is qualitatively different and the backward/forward asymmetry results in macroscopic directed motion of secondary electrons accompanying the passing of a fast electron beam through gas or plasma. The general formulas for this drag current are derived and applied to gaseous He.     

基于能力区域的交通状态预测方法

交通状态预测是交通流诱导和交通信息发布系统的重要依据.本文提出了一种基于能力区域的城市快速路交通状态预测方法,该方法通过构建神经网络分类器的能力区域,根据样本数据与交通状态类簇之间的空间距离,预测道路交通状态等级.神经网络分类器的能力区域能够有效融合时间、空间等多种特征,并且不需要考虑各特征之间的相关性,具有很强的适应性.实验结果表明,与经典的预测方法相比,其预测误差明显降低,均等系数增大,基于能力区域的方法预测交通状态具有较高的准确性.     

Dynamic characteristics of traffic flow with consideration of pedestrians’ road-crossing behavior

Pedestrians’ road-crossing behavior can often interrupt traffic flow and cause vehicle queueing. In this paper, we propose some moving rules for modeling the interaction of vehicles and pedestrians. The modified visual angle car-following model is presented for the movement of vehicles with consideration of the lateral effect of waiting pedestrians. The pedestrians’ behavior is summarized as consisting of three steps: pedestrian arrival, gap acceptance, and pedestrian crossing. Some characteristic parameters of pedestrians are introduced to characterize pedestrians’ behavior. Simulation results show that the interaction of vehicles and pedestrians lowers the traffic capacity and increases delays to both vehicles and pedestrians.     

光网络中视频业务传输的节点排队机制

对服务质量(QoS)中的队列机制进行了重点的分析, 并对先进先出(FIFO)以及具有优先级(PQ)的两种队列机制的特点进行了详细的比较。具有优先级(PQ)队列可以降低高优先级业务的平均排队时延, 但却无法保证时延抖动的要求。提出了基于具有优先级队列机制的改进方案。仿真结果表明, 新的排队机制有效得解决了具有优先级(PQ)排队机制中无法保证高优先级(视频)业务传输中的时延抖动问题, 当初始时延分布差别很大的数据包在经过节点排队调度之后, 时延抖动有了明显改善。     

Asymmetries in the nonmesonic weak decay of polarized lambda hypernuclei

The nonmesonic weak decay of polarized Lambda hypernuclei is studied for the first time by taking into account, with a Monte Carlo intranuclear cascade code, the nucleon final state interactions. A one-meson-exchange model is employed to describe the Lambda-->N-->nN processes in a finite nucleus framework. The relationship between the intrinsic Lambda asymmetry parameter a(Lambda) and the asymmetry a(M)(Lambda) accessible in experiments is discussed. A strong dependence of a(M)(Lambda) on nucleon final state interactions and detection threshold is obtained. Our results for a(M)(Lambda) are consistent with (11)(Lambda)B--> and (12)(Lambda)C--> data but disagree with observations in (5)(Lambda)H-->e.     

The influence of aggressive drivers on the properties of a stochastic traffic model

In this paper, we study numerically the influence of aggressive driving on the properties of the stochastic Nagel-Schreckenberg model, such as the traffic flow and the probability of car accidents. Hence, we find that these properties depend enormously on both the density and the fraction f of aggressive drivers. In addition, by studying the spatio-temporal organization of the vehicles, we show that at very low density, the traffic state transits from ordinary free traffic to queueing phase when we increase the fraction f. At relatively high density, the transition from congested traffic to queueing phase may also occur.Received: 15 April 2004, Published online: 21 October 2004PACS: 89.40.-a Transportation - 02.60.Cb Numerical simulation; solution of equations - 45.70.Vn Granular models of complex systems; traffic flow - 89.75.Fb Structures and organization in complex systems     

Influence of lateral discomfort on the stability of traffic flow based on visual angle car-following model

Due to the poor road markings and irregular driving behaviors, not every vehicle is positioned in the center of the lane. The deviation from the center can cause discomfort to drivers in the neighboring lane, which is referred to as lateral discomfort (or lateral friction). Such lateral discomfort can be incorporated into the driver stimulus–response framework by considering the visual angle and its changing rate from the psychological viewpoint. In this study, a two-lane visual angle based car-following model is proposed and its stability condition is obtained through linear stability theory. Further derivations indicate that the neutral stability line of the model is asymmetry and four factors including the vehicle width and length, the lateral separation and the sensitivity regarding the changing rate of visual angle have large impacts on the stability of traffic flow. Numerical simulations further verify these theoretical results, and demonstrate that the behaviors of diverging, merging and lane changing can break the original steady state and cause traffic fluctuations. However, these fluctuations may be alleviated to some extent by reducing the lateral discomfort.     

Final state effects of deep impurities in semiconductors

Current theoretical and computational models for optical excitation processes of deep level impurities in semiconductors tend to concentrate on the impurity itself, largely ignoring the effects the impurity has no final states. Using a spherical band model, we show these effects can be included in calculations on optical absorption processes. The band state is modified by the presence of the impurity and this modified state is used in the calculation of optical matrix elements. We show that final state effects can cause significant changes in the local density of states and optical matrix elements.     

Bivariate time-periodic Fokker-Planck model for freeway traffic

Recorded data of the density of cars and their speed from a German motorway are modeled by a bivariate Fokker-Planck equation. In order to cope with the evident diurnal variation, we assume a 24 h-periodicity in the drift and diffusion coefficients of this equation. After fitting these and smoothing them by polynomials, we validate the model by comparison of the empirical densities and densities generated by the model dynamics. We show that the time dependence of the drift field is related to a saddle-node bifurcation due to which the congested traffic state becomes stable. The separatrix between the basins of attraction is used to define flowing and jamming traffic during rush hours and characterizes the traffic dynamics together with the fixed points and the centre manifold.