基于局部线性嵌入和支持向量机回归的TBM施工参数预测

依托吉林引松工程开展隧道掘进机（TBM）施工参数预测研究,提出TBM施工数据分段提取算法,提取上升段前30 s的总推进力、刀盘转速、推进速度、刀盘扭矩、刀盘转速电位器设定值、推进速度电位器设定值、贯入度、贯入度指数(FPI)、扭矩切深指数(TPI)9个参数作为输入;通过局部线性嵌入（LLE）完成对上升段数据特征的降维;基于支持向量机回归（SVR）建立TBM施工控制参数（推进速度、刀盘转速）和负载参数（总推进力、刀盘扭矩）预测模型. 分析是否结合前一掘进循环的FPI、TPI指数进行预测对预测效果的影响. 结果表明,上述方法在推进速度、刀盘转速、总推进力、刀盘扭矩的预测中均取得了较好的预测效果,平均预测绝对百分比误差均小于15%,验证了该预测方法的有效性,该方法可以为TBM现场施工提供指导.     

实时人脸视频的色彩转换

针对视频捕捉时的人脸图像显示色彩感观差的问题，论述了图像间色彩转换的改进方法，并对人脸视频图像进行脸部和头发区域的分割，用改进后的色彩转换方法来改善实时视频的肤色和头发色彩，编写了VC环境下的算法实现程序，进行了大量的实验，实验结果表明此方法能实时转换人脸视频中每一帧的色彩，弥补了原视频显示方法的不足.     

基于ICC Profile的数码相机图像的颜色再现

基于国际色彩联盟制定的标准-ICC Profile文件来解决色彩在数码相机和显示器间传递时的一致性问题.借助于与设备无关的XYZ色彩空间,建立从数码相机色彩空间到显示器色彩空间之间的颜色转换关系,以实现数码相片色彩在显示器上的真实再现.利用数码相机和显示器的Profile,来检验数码相机和显示器之间的色彩再现效果,取得了满意的结果,证明了该方法的可行性.     

基于神经网络的摄像机色度标定方法

研究将摄像机颜色数据从设备依赖色空间转化到与设备无关色空间 .利用BP神经网络的非线性映射特性 ,实现RGB色空间到CIELAB色空间的映射 .给出用不同的网络结构得到的映射精度 .对于摄像机色空间的颜色数据转换 ,用BP网络可以获得令人满意的转换精度     

模糊神经网络在机载相机稳像中的应用

针对解决微型飞行器空中拍摄的图像抖动问题,采用自组织递归区间二型模糊神经网络的函数逼近及泛化能力对微型飞行器上的相机振动规律进行模拟,预测机载相机的振动矢量.该自组织递归区间二型模糊神经网络的初始规则数为零,所有规则都是通过结构和参数同时在线学习来产生,网络结构学习采用的是在线区间二型模糊群集,提高自组织递归区间二型模糊神经网络的稳定性及计算精度.仿真结果表明：将自组织递归区间二型模糊神经网络与双BP神经网络进行对比,利用自组织递归区间二型模糊神经网络对微型飞行器相机振动矢量进行预测的精度高.     

柔性翼微型飞行器试验样机研究

柔性翼微型飞行器是目前微型飞行器的重要发展方向，但是在总体、气动分析与实验等方面面临较大困难。通过对柔性翼微型飞行器试验样机的研制试飞和相关气动分析与试验，研究了解决柔性翼微型飞行器的气动性能、飞行性能与操纵稳定性的思路与方法。试验样机的研制试飞与相关风洞试验表明柔性翼微型飞行器具有飞行稳定，操纵反应适当和失速特性好等优点。     

Use of drilling performance to improve rock-breakage efficiencies: A part of mine-to-mill optimization studies in a hard-rock mine

In a hard-rock mine, blasting is an important rock-breakage process that impacts energy consumption both in downstream comminution processes and mine productivity. Optimizing the blast fragmentation to improve rock-breakage efficiencies during crushing and grinding is key to mine-to-mill(MTM) optimization. This study explores the use of monitoring while drilling(MWD) data to achieve this goal.Representative penetration rates(PRs) were extracted from blastholes to estimate intact rock properties and predict the breakage efficiencies that directly affect comminution energy consumption. Two intact rock properties, tensile strength(TS) and Bond work index(BWI), were correlated with the PR data to predict these efficiencies in crushing and grinding, respectively. Because of the complexity of the raw MWD data and effects of various disturbances, the MWD data was preprocessed and normalized to achieve a representative PR value at each blasthole. This preprocessing entailed defining valid PR ranges from the MWD data that could eliminate the noise related to discontinuity features in the rock mass structure as well as errors in operator behaviors. The PR data was also normalized using the adjusted penetration rate(APR) to minimize the effects of mechanical factors such as drill feed force, torque, and rotational speed. To correlate the representative APR value with intact rock properties, TS and BWI, various laboratory experiments were conducted: drilling tests using a high-precision coring machine, Brazilian disc tests, and Bond grindability tests. Based on the results of these experiments, models were developed to predict rock-breakage efficiencies during crushing and grinding based on APR. The result of this study can be used to obtain blast energy designs that consider comminution energy consumption and efficiency in the downstream rock-breakage processes.     

微型扑翼飞行器机翼气动特性研究

依据微型扑翼飞行器产生升力和推力的机理,设计了一套能够快速、有效求得扑翼飞行器机翼气动特性的计算方法。计算程序通过V isual Basic和Fortran语言混合编程来实现,核心部分是利用改进的片条理论方法估算扑翼机翼的气动性能。计算结果与在西北工业大学微型飞行器专用风洞中所进行的吹风试验结果吻合良好,证明了该方法的正确性和有效性。在此基础上,研究了不同机翼平面形状、不同展弦比、不同上下扑时间比对微型扑翼飞行器机翼气动性能的影响,这些参数对微型飞行器的设计有一定的指导和参考意义。     

基于分层MRF模型的抗抖动视频分割算法

提出了一种新的抗摄像机频繁抖动的视频分割算法.采用分层马尔可夫随机场（MRF）模型对视频各帧图像进行多分辨率建模，利用视频序列中帧图像的空间关系来提高分割的准确性，通过Gibbs采样算法求得最大后验概率（MAP），从而实现在摄像机抖动情况下对视频目标的准确分割.在强光、多目标以及复杂背景等情况下对视频序列的车辆目标进行分割.经过实验对比，新算法的分割效果明显优于背景累积相减分割算法以及高斯混合模型方法.     

基于Android的嵌入式视频智能车的软件设计

设计了一种通过Android移动设备来远程控制嵌入式视频智能车的软件系统,包括单片机软件和Android平台软件两部分.Android移动设备作为客户端通过WIFI方式与搭载有WIFI服务器的小车和网络摄像头通信,能完成控制车体行进、多角度实时画面监控、拍照和颜色识别等多项任务.测试表明,该系统功能完善可靠,适用于特殊环境下的视频监控领域.     

基于视频捕捉的小区闭路电视监控系统

提出了一种居住小区闭路电视监控系统设计方案 .该方案采用视频捕捉技术 ,用视频捕捉卡获取动态视频信号 ,对视频信号进行数字化处理并存储于计算机硬盘中 ,由设备驱动程序读取并进行显示 .系统利用楼宇可视对讲系统的CCD摄像机作为监控设备 ,使系统资源得以充分利用 .与传统的模拟式闭路电视监控系统相比 ,在图像的处理、显示与存储方面更加灵活 ,使用更加方便 ,并可节省价格昂贵的模拟信号存储设备和编辑处理设备 .所采用的视频捕捉卡支持黑白图像信号和彩色图像信号 ,便于系统升级     

一种弧焊机器人网络视频监控系统的设计与实现

为实现弧焊机器人的远程控制，基于DS-400xM系列图像采集卡开发了网络视频监控系统．其硬件由CCD摄像机、图像采集卡、视频服务器和客户端计算机构成，软件基于VC＋＋6．0开发．在实验室条件下对该系统进行了测试，结果表明，系统可行，可供被授权用户在网络中任何地方进行访问．     

土工离心模型试验中路堤分层填筑装置的研制

为了完善土工离心模型试验中的路堤填筑模拟技术,研制了一套在离心机运转条件下可编程控制的路堤分层填筑装置,其由箱式储砂结构、泄砂推拉机构、步进电机控制系统和摄像系统组成.采用步进电机-行星减速器-滚珠丝杠的组合方案实现大扭矩、高精度动力输出,使低摩阻、小变形的泄砂层与上部储砂结构前后错动,分隔于储砂结构内的高密度锆砂经承力底板上不等径、不等距泄砂孔定量下泄;通过在泄砂孔边缘加装折线形导流板,显著减弱离心场中Coriolis效应对下泄砂粒运动轨迹的影响,降落至地基面预设区的砂粒分层堆积为形状规整的路堤.试验表明,研制的填筑装置在60g离心加速度下,能模拟最小分层厚度20 mm的路堤填筑过程,模型横断面与设计面积相对误差为0.50%～9.50%,实现了模型与原型路堤在几何、密度和强度等方面的相似,具有系统可靠、结构紧凑、功能完整等技术特点.     

高速CCD视频摄像机的设计与研究

针对大视场、高帧频、多目标电视测量系统,介绍了高速CCD视频摄像机的设计和研究。研究高速CCD视频信号的产生,给出相应的电路设计。建立并分析了信号变换模型。基于设计的信号变换电路,可把高速CCD视频摄像机和标准TV监视器、录像机联合起来使用,实现高速CCD视频信号的显示与记录,而不需要昂贵的辅助仪器。实验结果表明高速CCD视频摄像机的适用性。     

船舶综合液压推进控制方法及仿真研究

为解决船舶在向大型化、高速化方向发展时船舶主机设计困难,机舱布置不合理以及机动性能差等问题,提出船舶综合液压推进,建立该系统的数学仿真模型,并通过试验台测试数据对该模型进行验证.针对该推进系统的特点,提出船舶综合液压推进恒功率与变功率联合控制方法(CVCC).基于配合元件的特性以及仿真模型,分别通过理论分析以及仿真试验对CVCC控制方法进行分析研究,两者相互印证,同时给出系统主要参数变化情况.结果表明,应用该控制方法可保证主机持续运行在最佳工况,使主机功率得以充分利用,克服了柴油机直接推进系统中,主机在轻载以及重载工况下无法发出设计功率的缺点.     

Experimental analysis on effect of grouser circumferential angle on motion performance of lunar rover wheel

To guide the design of grouser parameters for lunar rover wheel,the effect analysis of circumferential angle of grousers on motion performance of a rigid wheel was conducted in a laboratory single-wheel test bed with loose sand bin.All tests had been done at a free wheel sinkage.The circumferential angle was varied from 0°,5°,10°,15° to 20°.By the analysis and comparisons of tractive and steering performance between smooth wheels and wheels with different circumferential angle grousers,the results show that the drawbar pull(DP),driving torque(DT) and steering resistance moment(SRM) decrease slightly with the increasing circumferential angle.And chevron grousers are more beneficial to improve the tractive performance than herringbone grousers.In order to evaluate motion performance of wheels with grousers,the evaluation indexes of motion performance are introduced into the experimental results.The results reveal that the optimal circumferential angle is 0°,and the optimum value of wheel slip is 13% from the perspective of saving energy to the similar loose sand in the experiments.     

Numerical film torque modeling of friction pair in hydro-viscous clutchs

A numerical model to predict film torque of hydro-viscous clutch was developed.The model was established with computational fluid dynamics（CFD）.The pressure distribution,velocity of flow and film torque were obtained based on vertical-horizontal grooved plate and radial grooved plate separately.The boundary conditions,such as the relative rotation,the fluid temperature and the oil feeding pressure,were also discussed.The results showed that the film torque of two kinds of grooved plate increased with increasing relative rotation.However,the film torque decreased with increasing fluid temperature and feeding pressure.Meanwhile,the film torque of radial grooved plate was less than vertical-horizontal grooved plate at the same condition.Our study showed that the model can efficiently calculate the film torque with complex geometry parameters and boundary conditions.     

Dynamic torque response analysis of IPMSM in flux weakening region for HEV applications

An over-modulation based vector control strategy for interior permanent magnet synchronous machine (IPMSM) is proposed and investigated. The strategy increases the reference flux weakening voltage to improve efficiency in flux weakening region of IPMSM with the same dynamic torque response performance in standard SVM technique. The relationship between dynamic torque performance and the reference flux weakening voltage is also discussed. In order to achieve fast and smooth shift process, the torque response must be less than 20.ms in the parallel hybrid electric vehicle (HEV), according to this, modeling and experimental studies were carried out. The results show that the proposed strategy can achieve the same dynamic and steady state torque performance with higher reference flux weakening voltage, which means higher efficiency.     

不同迎角下前体涡流动的等离子体控制特性

应用一对单介质阻挡放电等离子体激励器对20°顶角圆锥-圆柱组合体圆锥段分离涡流场进行了主动控制研究。实验在3.0 m×1.6 m低速风洞中进行,迎角35°～70°,基于圆锥段底面直径的雷诺数为5.0×104。实验结果包括7个测量截面周向压力分布、由周向压力分布推断得到的截面处空间涡结构以及积分得到的截面当地力和圆锥段力。实验结果表明:(1)在35°～50°迎角范围内,圆锥段流场只有一对非对称的主涡,圆锥段分离涡流动呈现近似锥型流特性,随着迎角增大,圆锥段侧向力系数符号不变;(2)在50°～70°迎角范围内,圆锥段流场呈现多涡结构,圆锥段分离涡流动不再呈现锥型流特性,此时随着迎角增大,圆锥段侧向力系数会发生多次变号;(3)等离子体控制使得圆锥段对涡流场中第1个新涡出现的迎角推迟。     

Optimization algorithm for estimating the human pose by using the morphable model

An optimization algorithm is proposed utilizing the video data and point cloud data captured by the depth camera to solve the problems such as error-proneness and incoherence of motion sequence caused by the existing human pose estimation algorithms based on the morphable model. For video data, the neural network is first used in extracting the model parameters from each color image frame. Next, the human key-points and contour constraint are considered to optimize the above parameters. Then the coherence between every two consecutive frames is utilized to correct the error of pose estimation, thus making the resulting motion sequence smoother. In addition, the point cloud and the model obtained from the corresponding color image frame are used as the joint input to further improve the estimation accuracy. Finally, the distance between the point cloud and the corresponding point of the model is constrained to be as small as possible to obtain a more reasonable solution. The proposed algorithm and the state-of-the-art algorithms are compared qualitatively and quantitatively on the data set and real video set. Experimental results show that the algorithm can effectively correct the error and incoherence in the single-frame pose estimation results and greatly improve the accuracy when using point cloud data optimization.