两相流流型多尺度熵及动力学特性分析

研究了几种典型非线性时间序列的多尺度熵特征,在此基础上分析了由插入式阵列电导传感器采集的144种流动条件下的垂直上升气液两相流电导波动信号.研究结果表明：利用小尺度下样本熵的变化速率特征可以分辨三种典型流型（泡状流、段塞流、混状流）,而大尺度下样本熵的波动特征可以反映各种流型的动力学特性.泡状流随机可变特性表现为大尺度下样本熵的高值及振荡特征;段塞流气塞与液塞的间歇性运动表现为大尺度下样本熵的低值及平稳性;混状流极不稳定的振荡运动特性表现为介于泡状流及段塞流之间的熵值特点,并在更大尺度时熵值逐渐接近泡状流 关键词： 样本熵 多尺度熵 气液两相流 动力学特性     

多元时间序列复杂网络流型动力学分析

针对气液两相流流动特性,利用有限元分析方法设计变曲率对壁式电导传感器.采用设计加工的传感器在多相流装置上进行气液两相流动态实验,并测得多组对应于不同流型的电导波动信号. 基于测量数据,采用多元时间序列复杂网络构建算法构建对应于不同流型的复杂网络.在此基础上, 对网络的社团特性进行了分析, 研究发现,不同的社团结构对应于不同的流型,而社团内部网络特征可有效刻画不同流型内在动力学特性.多元时间序列复杂网络分析可为两相流流型演化动力学特性研究及流型识别提供新理论、开拓新途经.     

基于混沌吸引子形态特性的两相流流型分类方法研究

利用高灵敏度差压传感器，在垂直上升管中采集到了80组气液两相流差压波动信号.利用放置参考截面的方法，建立了描述混沌吸引子形态的一般方法，在此基础上，提出了不同维数下的吸引子形态特征量进行组合的气液两相流流型分类新方法.研究结果表明：该方法对包括复杂过渡流型在内的气液两相流流型有很好分类效果，预示着混沌吸引子形态描述是研究非线性时间序列的实用有效途径. 关键词： 气液两相流 流型分类 吸引子形态 混合维     

螺杆螺旋通道内气液两相流流动特性研究

本文以水和空气为工质对螺杆螺旋通道内气液两相流流动特性进行研究,采用可视化的方法对螺杆螺旋通道内流型结构进行观察,观察到通道内不同转角位置处的气液两相流型及其特征。根据实验观测结果,绘制了三个转角位置的流型图,并对各流型特征和流型之间的转换界限进行分析,给出了泡状流向间歇流转变、环状流向间歇流转变的准则关系式。     

两相流流型动力学特征多尺度递归定量分析

基于垂直上升管中测取的气液两相流电导波动信号,采用递归定量分析方法,从多尺度角度研究了气液两相流泡状流、段塞流及混状流三种典型流型的动力学运动特征.研究结果表明,低频泡状流及混状流在递归图表现为沿对角线方向比较发育的混沌递归线条纹理特征,表明了低频运动的泡状流及混状流具有较好的确定性运动行为,而随着泡状流及混状流运动频率增加,混沌递归特征变差,其运动特征逐渐向随机方向发展.对于段塞流,在混沌递归图上逐渐呈现间歇的矩形块纹理结构,且段塞流中液塞与气塞的间歇运动特征出现在高频段,而段塞流中的泡状流运动则出现在低频段上,且随着泡状流运动频率增加,泡状流逐渐失去确定性运动行为.表明了基于电导波动信号的多尺度非线性分析方法是理解与表征气液两相流动力学特性的有效途径. 关键词： 两相流 流动特性 多尺度分析 递归分析     

基于连续图像灰度序列混沌特性的油气水三相流型识别

以35号润滑油、空气和自来水为试验介质,应用高速摄像机对垂直上升管内的油气水三相流的六种典型流型进行了动态图像的拍摄.提取每一帧图像的灰度均值组成时间序列,对其进行混沌特性分析,提取序列的HURST指数,关联维以及分别以2和e为底的最大李亚普诺夫指数,组成特征向量,输入支持向量机进行流型分类.试验结果表明:连续图像的灰度时间序列的混沌特性能够对油气水三相流的典型流型进行很好的表征,结合支持向量机进行分类,识别率达到90%以上,为流型的在线识别提供了一种新方法.     

两相流相空间多元图重心轨迹动力学特征

本文提出了一种新的混沌时间序列高维相空间多元图重心轨迹动力学特征提取方法. 在确定了最佳嵌入维数和延迟时间后, 将相空间中高维矢量点映射到二维平面的雷达图上, 相应地将相空间中高维矢量点变换为对应的几何多边形. 通过提取几何多边形的重心位置得到重心轨迹动力学演化特性, 并利用重心轨迹矩特征量区分不同性质的混沌时间序列. 在此基础上, 处理分析了气液两相流电导传感器动态信号, 发现高维相空间多元图重心轨迹矩特征量不仅可以辨识泡状流、段塞流和混状流, 而且为流型动力学演化机理提供了新的分析途径.     

气液两相流电导波动信号复杂性测度分析及其流型表征

为了考察从时间序列提取的复杂性测度与气液两相流流型变化之间的关系，本文首先讨论了三种复杂性测度(Lempel-Ziv复杂性、功率谱熵和近似熵)对周期信号、随机信号、混合随机信号和混沌信号的识别能力，然后分析了时间序列长度对复杂性计算的影响.在此基础上，从实际测量的80种垂直上升管中气液两相流电导波动信号中提取了这三种复杂性测度，结果表明：三种复杂度对两相流流型变化是敏感的，通过对三种复杂度随两相流流动参数变化规律分析，可以得到气液两相流动力学结构反演特征，为揭示气液两相流流型转化机理提供了一种有效的辅助诊断工具. 关键词： 气液两相流 Lempel和Ziv复杂性 功率谱熵 近似熵     

正方形小通道内的非牛顿流体-氮气两相流动流型的实验研究

对水力直径为2.5 mm的正方形小通道内的非牛顿流体-氮气的垂直向上两相流动流型进行了可视化实验,工质分别为:浓度0.2%的聚丙烯酰胺(PAM)和0.2%的黄原胶(XG)水溶液,表观气速0.1～100 m/s,表观液速0.01～6 m/s.观察到的典型流型有:弹状流、搅拌流、弹环状流和环状流,其中弹环状流未见于水-空气上升流动.在PAM-氮气实验中发现了一种新流型-泡状-弹状流.通过流型图对比,发现非牛顿流体的搅拌流区域较牛顿流体窄,弹状-搅拌流转变线也明显右移,非牛顿流体的黏性对流型转变的影响较大.     

垂直气液两相流混沌吸引子单元面积分析

为了充分反映吸引子结构随时间延迟的变化规律,在现有吸引子形态描述方法基础上定义了吸引子单元面积,通过仿真发现,吸引子单元面积随时间延迟变化曲线第一个波峰的高度和时间延迟主要由信号中大幅值波动的数量、频率决定,利用此规律对实验采集到的气液两相流电导波动信号进行分析,发现在固定液相流量条件下,改变气相流量会导致泡状流、段塞流和混状流中大幅值波动幅度的改变,但相同流型信号中大幅值波动的频率比较接近.将吸引子单元面积随时间延迟变化曲线第一个波峰的时间延迟和落差比作为特征量,可以实现泡状流、段塞流、混状流的流型分类.     

小通道氮气-水两相流三谱切片波动特性及其流型表征

将三谱切片与波动特性分析方法相结合, 组建了一种新的非线性分析方法, 并讨论了时间序列长度与快速傅里叶变换(FFT)长度对三谱切片的波动特征值的影响, 发现在无噪声干扰情况下, 序列长度与FFT 长度与矢量类间距值成正比, 在有噪声干扰情况下, 近似成反比关系.将该分析方法应用到分形序列(Brown)、混沌序列(Lorenz)和周期序列(正弦), 检验了波动特性方法的抗噪能力和表征复杂度能力. 结果表明: 该波动特性分析方法较其他功率谱分析方法对分形和混沌序列具有较好的抗噪能力, 而对周期序列的抗噪能力相对较弱; 该波动特性方法对序列内在复杂度的表征具有较好的效果.在此基础上, 对水力直径为1.15 mm的矩形小通道(w×h=2 mm×0.81 mm)中的 矩形通道内氮气-水两相流型的差压信号进行了研究. 通过对流型差压信号三谱切片的分析揭示了不同流型的主振荡模式二次耦合现象, 提取不同流型三谱切片的波动特征值, 实现了小通道氮气-水两相流典型流型的准确识别, 同时也为其他不同介质的多相流动特性分析与流型辨识提供了一个新的思路. 关键词： 三谱切片 波动特性 流型识别 流型动力学     

分数阶Willis环脑迟发性动脉瘤时滞系统混沌分析

在脑血动脉瘤的临床研究中,Willis环脑动脉血管瘤系统(Willis aneurysm system,WAS)起着重要作用,分数阶WAS尽管能进一步加深该系统的机理刻画,但是不能描述原因不明的迟发性动脉瘤.鉴于此,本文提出分数阶Willis环脑迟发性动脉瘤时滞系统(fractional Willis aneurysm system with time-delay,FWASTD)并验证了其有效性;利用时间序列图、相图、Poincaré截面等证实了FWASTD的混沌特性;研究时滞对于系统的重要生理参量的影响,发现了血流阻力系数在时滞状态下对系统稳定的重要性;根据分数阶时滞系统的稳定性理论,设计相应线性控制器,对FWASTD进行了有效控制,同时也探讨了时滞系统的自同步控制.本文完善了脑动脉瘤系统的理论基础.     

Theoretical vs. empirical classification and prediction of congested traffic states

Starting from the instability diagram of a traffic flow model, we derive conditions for the occurrence of congested traffic states, their appearance, their spreading in space and time, and the related increase in travel times. We discuss the terminology of traffic phases and give empirical evidence for the existence of a phase diagram of traffic states. In contrast to previously presented phase diagrams, it is shown that “widening synchronized patterns” are possible, if the maximum flow is located inside of a metastable density regime. Moreover, for various kinds of traffic models with different instability diagrams it is discussed, how the related phase diagrams are expected to approximately look like. Apart from this, it is pointed out that combinations of on- and off-ramps create different patterns than a single, isolated on-ramp.     

Synchronized Flow in a Cellular Automaton Model with Time Headway Dependent Randomization

We propose another possible mechanism of synchronized flow, i,e. that a time headway dependent randomization can exhibit synchronized flow. Based on this assumption, we present a new cellular automaton （CA） model for traffic flow, in which randomization effect is enhanced with the decrease of time headway. We study fundamental diagram and spatial-temporal diagrams of the model and perform microscopic analysis of time series data, which shows the model could reproduce synchronized flow as expected. It is also shown that a spontaneous transition from synchronized flow to jam could be observed by incorporating slow-to-start effect into the model. We expect that our work could contribute to the understanding of the real origin of synchronized flow.     

Traffic states and fundamental diagram in cellular automaton model of vehicular traffic controlled by signals

We present a cellular automaton (CA) model for vehicular traffic controlled by traffic lights. The CA model is not described by a set of rules, but is given by a simple difference equation. The vehicular motion varies highly with both signals’ characteristics and vehicular density. The dependence of tour time on both cycle time and vehicular density is clarified. In the dilute limit of vehicles, the vehicular motion is compared with that by the nonlinear-map model. The fundamental diagrams are derived numerically. It is shown that the fundamental diagram depends highly on the signals’ characteristics. The traffic states are shown for various values of cycle time in the fundamental diagram. We also study the effect of a slow vehicle on the traffic flow.     

The influence of nonmonotonic synchronized flow branch in a cellular automaton traffic flow model

In this paper we study the congested patterns upstream of an isolated on-ramp in a cellular automaton traffic flow model, which is proposed in our previous paper [Cheng-Jie Jin, Wei Wang, Rui Jiang, Kun Gao, J. Stat. Mech (2010) P03018]. The simulation results under open boundary conditions are presented by spatiotemporal diagrams. Our diagram of congested patterns is quite similar to that of the cellular automaton models within Kerner’s three-phase traffic theory, while some differences in the “moving synchronized flow pattern” (MSP) should be noted. In our model the upstream front of MSP propagates not only upstream, but also downstream. The propagation direction depends on the flow rates and densities of free flow and synchronized flow. Besides, in our model the outflow of wide moving jams or bottlenecks could be free flow or synchronized flow, as reported in many empirical data. In the dissolving of congestions, the form of free flow may be hindered and stable synchronized flow may emerge. This phenomenon can help us understand more about the outflow. All the interesting characteristics of our model are due to the nonmonotonic structure of synchronized flow branch in the fundamental diagram, which has not been found in previous models.     

Effect of nonlinear translations on the pulsation of cavitation bubbles

The pulsations and translations of cavitation bubbles obey combined ordinary differential equations, and their nonlinearities are studied by the bifurcation diagram and the phase diagram in a strong ultrasonic field. Bubble pulsation can change regularly or irregularly with changing driving pressure in the time domain. The bifurcation diagrams of the pulsation versus driving pressure show that the pulsations and translations of bubbles have nonlinear characteristics, and the nonlinear translations of bubbles can disorder the pulsations for certain parameters. Disorder of the pulsation can also be caused by nonlinear pulsation itself. In addition, the phase diagrams also show that the nonlinear translations make a large contribution to the pulsations. The same result can also be obtained when the ambient radii of two bubbles are different.     

混有协同自适应巡航控制车辆的异质交通流稳定性解析与基本图模型

针对传统车辆和协同自适应巡航控制(cooperative adaptive cruise control,CACC)车辆构成的异质交通流,研究其稳定性与基本图模型.应用实车测试验证的CACC模型和智能驾驶员模型(intelligent driver model)分别作为CACC车辆和传统车辆的跟驰模型,建立异质流稳定性解析框架,研究不同平衡态速度、不同CACC车辆比例时的异质流稳定性.推导异质流基本图模型,并进行数值仿真实验.研究结果表明,在传统车辆稳定的速度范围,异质流处于稳定状态.在传统车辆不稳定的速度范围,CACC车辆比例增加以及平衡态速度远离9.6—18.6 m/s速度范围,均能够改善异质流的不稳定性.通行能力随着CACC车辆比例的增加而提高.此外,CACC模型的期望车间时距越大,异质流稳定域越大,但通行能力降低.因此,恒定车间时距CACC控制策略下的期望车间时距取值应权衡异质流稳定域和通行能力两个方面的影响.     

Synchronized traffic flow simulating with cellular automata model

The synchronized flow traffic phase of Kerner’s three-phase traffic theory can be well reproduced by the model proposed by Jiang and Wu [R. Jiang, Q.S. Wu, J. Phys. A: Math. Gen. 36 (2003) 381]. But in the Jiang and Wu model, the rule for brake light-after switching on, the brake light will not set off until the vehicle accelerates-is obviously unrealistic. Thus we improved the model by considering the difference in accelerating and decelerating performance under different driving conditions. The fundamental diagram and spatial-temporal diagrams are analyzed. We confirmed that the new model could reproduce the synchronized flow by two methods, i.e. the traffic flow interruption effect and performing microscopic analysis of time series data. Simulation results show that the decelerating difference is an important factor to reproduce the synchronized flow. We expect that our work could make contributions to understanding the mechanism of the synchronized flow.     

基于PNN网络和Freeman链码的抽油机井工况诊断

利用概率神经网络(PNN)对抽油机井工况进行诊断,建立了抽油机井工况诊断的概率神经网络模型。对示功图提取特征值的质量好坏直接影响识别效率和可靠性,提出了用Freeman链码对等效的电流示功图提取特征参数,进行预处理,建立抽油机典型工况的链码特征样本库。将Freeman链码作为特征向量,利用MATLAB对网络进行训练。结果表明,Freeman链码能够有效的识别各种典型工况示功图,并且该概率神经网络学习速度快、诊断准确率高,可用于抽油机井工况的实时监测和诊断。