微尺度射流流量计的设计与仿真研究

MEMS流量控制系统中,需要微型化的流量传感器,设计了一种反馈流驱动的微尺度射流流量计.采用数值模拟的方法对微尺度射流流量计进行研究,CFD软件的仿真结果表明:在较宽入口速度范围内该微尺度射流流量计起振快,振荡稳定,切换灵敏,斯特劳哈尔数基本恒定,并且流速测量下限可以达到0.075 m/s,在一定程度上为射流流量计的微...     

A cross-junction channel valveless-micropump with PZT actuation

A gas-jet micro pump with novel cross-junction channel has been designed and fabricated using a Si micromachining process. The valveless micro pump is composed of a piezoelectric lead zirconate titanate (PZT) diaphragm actuator and fluidic network. The design of the valveless pump focuses on a cross-junction formed by the neck of the pump chamber and one outlet and two opposite inlet channels. The structure of cross-junction allows differences in fluidic resistance and fluidic momentum inside the channels during each PZT diaphragm vibration cycle, which leads to the gas flow being rectified without valves. The flow channels were easily fabricated by using silicon etching process. To investigate the effects of the structure of the cross-junction on the gas flow rate, two types of pump with different cross-junction were studied. The design and simulation were done using ANSYS-Fluent software. The simulations and experimental data revealed that the step-nozzle structure is much more advantageous than the planar structure. A flow rate of 5.2 ml/min was obtained for the pump with step structure when the pump was driven at its resonant frequency of 7.9 kHz by a sinusoidal voltage of 50 V_p–p.     

Fabrication of Robust 2-D and 3-D Microfluidic Networks for Lab-on-a-Chip Bioassays

We describe a new method for the fabrication of robust three-dimensional (3-D) lab-on-a-chip devices comprising microfluidic channels and integrated electroanalytical sensors and show their potential application in supporting leak-free pressure driven fluid flow. The technology is based upon the lithographic patterning of a soluble negative photoresist and an insoluble, crosslinked, positive photoresist (SU-8) to produce the pattern of a large area microfluidic network, followed by coating the pattern with silicone rubber or PDMS. After curing, immersion in acetone rapidly elutes the soluble resist to yield the network of channels. As a result the channels formed are robust and it is possible to fabricate a device that supports fluids flowing at high flow rates (without leakage) without the need for special pre-treatment or surface modification of the substrate or silicone covering. We also demonstrate the flexibility of the methodology by producing more complex fluidic networks that comprise bridges which enable one fluid stream to pass over (and remain physically separated from) a second fluid stream. The sealing and packaging techniques, developed for this process do not require the use of extreme conditions and are thus compatible with a broad range of lithographic processes (including, for example, the definition of and incorporation of in situ electrochemical sensor elements). Finally, measurements on a model bioelectroanalytical system applicable to analysis of a model immunoassay are shown, demonstrating the application of this new fabrication method in developing reproducible analytical methods within an encapsulated device.hfillhbox[1297]     

基于神经网络的城市轨道交通短期客流预测研究

该文结合神经网络来研究城市轨道交通中短期客流预测问题。设计出了基于自回归神经网络的轨道交通客流预测模型、模型描述及其模型训练算法。通过matlab仿真实验来验证预测模型的性能,优于将最小二乘支持向量机与离散一维Daub4小波分析结合起来预测效果。     

基于混沌RBF神经网络的汽油机进气流量预测研究

针对汽油机进气流量的多维非线性特性,提出了一种混沌径向基（RBF）神经网络的汽油机进气流量预测模型。证明了汽油机进气流量时间序列具有混沌特性,对采集的原始数据进行相空间重构,利用RBF神经网络对重构后的数据进行训练和预测,并利用混沌算法确定输出层连接权值和隐含层高斯函数径向基中心,使其达到全局最优,加快了RBF神经网络的收敛速度。仿真结果表明,与空气进气流量平均值模型、RBF神经网络预测模型比较,该模型具有更高的预测精度,为精确及时测试汽油机进气流量提供了一种全新的软件测量方法。     

基于SNMP的P2P网络流量监控模型

该文提出了一种基于SNMP的P2P网络流量监控模型。该模型借鉴了SNMP思想和ECGP层次结构模型,在P2P网络中设置监控管理站点,并通过逻辑树型结构组织起来,以达到对整个网络流量进行监控并对超负荷节点路由进行合理调整的目的。该文实验采用C 进行系统模拟,实验结果表明该方法对调整网络流量具有十分明显的效果。因此在目前P2P网络中,具有一定的推广和使用价值。     

CS算法优化BP神经网络的短时交通流量预测

为了提高短时交通流量的预测精度,提出一种布谷鸟搜索算法优化BP神经网络参数的短时交通流量预测模型（CS-BPNN）。基于混沌理论对短时交通流量时间序列进行相空间重构,将重构后的时间序列输入到BP神经网络进行学习,采用布谷鸟搜索算法找到BP神经网络最优参数,建立短时交通流量预测模型,通过具体实例对CS-BPNN性能进行测试。仿真结果表明,相对于对比模型,CS-BPNN提高了短时交通流量的预测精度,更加准确反映了短时交通流量的变化趋势。     

城市交通网络宏观控制模型及仿真实现

为研究适合城市交通网络控制系统应用的交通流预测模型,在改进Van Den Berg, M的路段交通流模型的基础上,建立了以路口交通流为基本建模单元,以动态非线性离散方程反映交通流变化的城市交通网络宏观模型.为验证该模型能有效地预测城市路网的交通流信息,在VC++net环境下,开发了城市交通宏观控制模型仿真系统UTFS,设计了网络拓扑结构模块,以适应不同规模、不同复杂程度的实际交通网络的仿真要求.最后选取典型网络进行应用研究.仿真结果表明:该模型满足交通控制对控制模型的实时性和精度要求,该仿真系统可以作为城市交通网络宏观控制模型验证的有效工具,也可以作为城市交通控制系统控制和优化研究的辅助工具.     

Computational and performance analysis of a continuous magnetophoretic bioseparation chip with alternating magnetic fields

This paper presents the modeling and optimization of a magnetophoretic bioseparation chip for isolating cells, such as circulating tumor cells from the peripheral blood. The chip consists of a continuous-flow microfluidic platform that contains locally engineered magnetic field gradients. The high-gradient magnetic field produced by the magnets is spatially non-uniform and gives rise to an attractive force on magnetic particles flowing through a fluidic channel. Simulations of the particle–fluid transport and the magnetic force are performed to predict the trajectories and capture lengths of the particles within the fluidic channel. The computational model takes into account key forces, such as the magnetic and fluidic forces and their effect on design parameters for an effective separation. The results show that the microfluidic device has the capability of separating various cells from their native environment. An experimental study is also conducted to verify and validate the simulation results. Finally, to improve the performance of the separation device, a parametric study is performed to investigate the effects of the magnetic bead size, cell size, number of beads per cell, and flow rate on the cell separation performance.     

Pneumatically bidirectional microfluidic regulation using Venturi pumps by deep RIE and bonding technology

 A novel design for bidirectional fluidic motion has been demonstrated which is widely used in the biochip or microfluidic component. Two miniaturized Venturi pumps as well as pneumatic servo system are designed to easily control the bidirectional fluidic motion by simple fabrication. The pumping velocity is 0.86 μl/min at a 2.75 slpm (standard liter per minute) air flow read from mass flow controller (MFC) for totally 4.3 μl blue ink in a 300 μm wide by 300 μm deep channel. The higher airflow, the faster fluidic pumping speed. Numerical simulation is performed to correlate the experimental data of fluidic speed and air flow in microchannel. The test chip with two Venturi pumps and channel was batchedly fabricated by silicon deep reactive ion etching (RIE) and glass anodic bonding. The ICP LIGA process is also investigated after deep RIE followed the electroforming and hot embossing. Received: 10 August 2001/Accepted: 24 September 2001     

Combined circuit/device modeling and simulation of integrated microfluidic systems

A combined circuit/device model for the analysis of integrated microfluidic systems is presented. The complete model of an integrated microfluidic device incorporates modeling of fluidic transport, chemical reaction, reagent mixing, and separation. The fluidic flow is generated by an applied electrical field or by a combined electrical field and pressure gradient. In the proposed circuit/device model, the fluidic network has been represented by a circuit model and the functional units of the /spl mu/-TAS (micro Total Analysis System) have been represented by appropriate device models. We demonstrate the integration of the circuit and the device models by using an example, where the output from the fluidic transport module serves as the input for the other modules such as mixing, chemical reaction and separation. The combined circuit/device model can be used for analysis and design of entire microfluidic systems with very little computational expense, while maintaining the desired level of accuracy.     

微小尺度射流流量传感器的设计与仿真分析

在许多新型传感器和微系统中均存在微量流体自动、精确地驱动和控制问题,而这有赖于微小尺度下对流量的精确测量。基于射流振荡原理设计了一种新型的无反馈通道微小尺度流量传感器,采用计算流体动力学(CFD)方法对该流量传感器的测量特性进行了仿真研究。通过观测振荡腔内部流场,分析了振荡器内部流动形态和射流振荡过程。通过对监测点压力变化曲线的分析,获得了不同入口速度下流体振荡频率,建立了流体流速与振荡频率的函数关系。研究结果表明,该微小尺度射流振荡器振荡平稳,主射流切换灵活,在较宽的流速范围内,流速与振荡频率具有线性关系,具有0.3%的较低的相对压力损失并可达到较小的测量下限,易于加工成型。     

径向基网络在密相气力输送中的应用

在水平管道中,用压缩空气对玻璃珠进行密相气力输送压损实验研究。管道压降是密相气力输送系统的关键参数之一,它的大小基本决定了动力的大小。传统的方法完全靠经验公式和经验来决定动力大小,误差较大。本文提出了基于径向基网络的密相气力输送管道压降模型,对不同流态下管道压降进行了仿真。结果表明,RBF网络能对不同流态下的管道压降进行较好的仿真。RBF网络的收敛速度快,可实现密相气力输送参数的在线控制。     

基于ResNet的稳健语音DOA估计算法

针对传统波达方向（Direction of arrival, DOA）估计算法在阵列模型误差条件下估计性能下降的问题,提出了一种基于残差网络（Residual network, ResNet）的DOA估计算法。该算法根据神经网络数据驱动不依赖阵列流型的特点,从广义互相关（Generalized cross-correlation, GCC）中提取特征,将提取的特征作为神经网络深层分类器的输入,对信号分类;根据分类结果选取对应子区间数据进行训练,建立ResNet学习特征与DOA估计之间的非线性映射关系,形成一个数据驱动的稳健DOA估计系统。仿真与实验验证结果表明,该算法有效地解决了传统DOA算法在阵列模型误差条件下无法准确得到DOA结果的问题。     

微机械开放式射流陀螺的敏感原理和数值仿真

提出了一种基于开放式气流通道的微机械射流陀螺的设计、敏感原理和数值仿真.敏感元件内部气流的运动采用基于流固耦合的三维瞬态数值分析计算,计算和测试结果表明:①气流在压电泵的驱动下,由开放式气流通道的两个入口进入并汇聚,由喷口喷出并在射流敏感室形成射流敏感体,在入口和出口之间形成定向流动.②射流敏感体在哥氏力的作用下发生偏转,与作为电桥臂的热线发生热量交换,导致电桥输出与输入角速度成正比的不平衡电压.③陀螺灵敏度为SFZ=2.0μV/°/s,非线性度优于0.5%.     

面向复杂网络存储系统的元胞自动机动力学分析方法

大规模网络存储系统中复杂的数据传输行为隐藏着一定的动力学规律性.针对基于对象的大规模网络存储系统,结合存储对象的智能性和主动性特征,分别在宏观与微观两个层次上提出了用于复杂网络存储动态行为规律分析的存储元胞自动机模型SNCA和OSDCA.在SNCA模型中,对网格拓扑结构的存储网络,结合存储对象的生命周期属性,可在宏观上分析网络存储系统的数据流动规律,确定存储网络拥塞程度,仿真结果揭示数据对象流动和存储网络中的相变具有全局相关性;在OSDCA模型中,综合热点数据的迁移和复制机制,在微观上分析I/O负载动态分布特性和存储热点迁移规律,仿真结果表明对象存储系统中的数据分布具有一定的自组织特性.     

基于LSTM-BP组合模型的短时交通流预测

为减轻日益严重的交通拥堵问题,实现智能交通管控,给交通流诱导和交通出行提供准确实时的交通流预测数据,设计了基于长短时记忆神经网络（LSTM）和BP神经网络结合的LSTM-BP组合模型算法.挖掘已知交通流数据的特征因子,建立时间序列预测模型框架,借助Matlab完成从数据的处理到模型的仿真,实现基于LSTM-BP的短时交通流精确预测.通过与LSTM\BP\WNN三种预测网络模型的对比,结果表明LSTM-BP预测的时间序列具有较高的精度和稳定性.该模型的搭建,可对交通分布的预测、交通方式的划分、实时交通流的分配提供依据和参考.     

Development of a MEMS microvalve array for fluid flow control

A microelectromechanical system (MEMS) microvalve array for fluid flow control is described. The device consists of a parallel array of surface-micromachined binary microvalves working cooperatively to achieve precision how control on a macroscopic level. Flow rate across the microvalve array is proportional to the number of microvalves open, yielding a scalable high-precision fluidic control system. Device design and fabrication, using a one-level polycrystalline silicon surface-micromachining process combined with a single anisotropic bulk etching process are detailed. Performance measurements on fabricated devices confirm feasibility of the fluidic control concept and robustness of the electromechanical design. Air-flow rates of 150 ml/min for a pressure differential of 10 kPa were demonstrated. Linear flow control was achieved over a wide range of operating flow rates. A continuum fluidic model based on incompressible low Reynolds number flow theory was implemented using a finite-difference approximation. The model accurately predicted the effect of microvalve diaphragm compliance on flow rate. Excellent agreement between theoretical predictions and experimental data was obtained over the entire range of flow conditions tested experimentally     

基于复杂网络的路网交通拥堵评估仿真模型

为了能够深入分析路网交通拥堵动态演进过程, 为交通拥堵治理提供决策分析工具, 提出了一个基于复杂网络的路网拥堵评估仿真模型。通过将路段阻抗概念引入复杂网络理论, 实现了路网拓扑模型和流量模型的结合; 此外, 还提出了虚拟测试车辆遍历的路网通行能力评估方法, 对不同路网条件下的道路通行能力进行评估。最后在PC系统上实现了完整的路网拥堵评估仿真系统, 并利用该系统对拥堵路段数量、交通流量和路网拓扑结构等因素对路网通行能力的影响进行了仿真分析。仿真实验结果与经典交通理论分析结果一致, 并能反映更多的动态过程信息, 表明该模型能够准确有效地进行交通路网拥堵评估, 可以为相应的交通拥堵管理决策提供依据。     

基于O-D矩阵估计的路网交通流量仿真模型

为了将交通出行需求对路网交通流量的影响进行动态的量化分析,提出了一个基于O-D矩阵估计的路网交通流量仿真模型。利用O-D矩阵估计的重力模型计算方法、复杂网络理论和路段阻抗模型,构建了路网模型;在人们出行总是选择路段阻抗最小路径的假定下,设计了出行需求的路网流量映射算法;基于离散事件仿真,在PC系统上实现了路网流量仿真系统。仿真结果表明：该仿真系统可以根据各交通子区域出行需求的变化,精确模拟路网流量和交通状态的动态演进。