Detection of crowdedness in bus compartments based on ResNet algorithm and video images

Multimedia Tools and Applications - The crowding in bus is an important factor affecting passenger satisfaction and bus dispatching level. However, how to use video images to detect crowding...     

基于多学科协同的某锁机构系统故障仿真

为解决机电液耦合的复杂运动机构系统的故障仿真问题,提出一种基于多学科软件协同的故障仿真分析方法.通过软件接口建立机构系统基于多学科协同的仿真模型.将相关故障模式影响因素注入机构系统仿真模型,仿真分析其对系统参数的影响.某锁机构系统的上锁故障仿真分析表明,该方法对工程实际复杂运动机构系统有效可行.该故障仿真方法有助于提高机械产品可靠性,也可为复杂机械系统的故障机理分析、故障诊断和维修等提供数据支持。     

二甲基氨基甲叉二膦酸氮氧化物的合成

以二甲基氨基甲叉二膦酸(DMMP)为原料,双氧水为氧化剂合成二甲基氨基甲叉二膦酸氮氧化物(DMMPO)。研究了原料配比、温度、时间及合成时pH值等条件对目标产物阻磷酸钙垢效果的影响,结果表明:n(DMMP)∶n(H2O2)=1∶3、反应温度为75℃、反应时间为2.5 h,并控制pH值为9左右时合成的DMMPO静态阻垢效果最好。通过IR和31P-NMR对产物进行了结构鉴定,同时对产品进行了活性物含量、亚磷酸根和磷酸根含量的测定,直接合成产物中活性组分含量为22.7%,亚磷酸根和磷酸根含量均小于0.1%。     

制粉工艺对贮氢合金电化学性能的影响研究

研究了干法、湿法和湿法加添加剂三种制粉工艺对贮氢合金电化学性能的影响 ,结果表明 :制粉工艺对合金粉末的表面形貌、粒度分布、氧含量均存在很大的影响 ,并影响到其电化学性能。当采用湿法制粉工艺时 ,贮氢合金电极的活化性能较干法制粉明显改善 ,若加入一定量的添加剂 ,活化性能进一步提高 ,第一次放电容量即达到2 78mAh·g-1,第二次放电容量达到合金电极的稳定容量 32 2mAh·g-1。湿法制粉对贮氢合金电极的循环稳定性能有一定的负面影响 ,这可能与贮氢合金粉末表面的氧化和细粉相对含量较多有关 ,然而 ,当加入一定量的添加剂以后 ,贮氢合金电极的氧含量得到有效的控制 ,3C充放循环稳定寿命达到 5 0 0次 ,与干法制粉工艺接近。     

基于高斯过程动态模型的人体节奏运动合成

提出一种新的基于高斯过程动态模型的节奏转移方法。该方法能够准确、有效地将现有运动中的节奏信息转移到新的运动中去,适用于各种不同类型的运动。首先,使用短时（short term）PCA计算源运动的节奏点,组合3种重要的运动特征求解目标运动的特征点;然后,使用动态规划算法来找到两者之间的最佳匹配,最大化减少计算时间及对目标运动的修改;最后,使用高斯过程动态模型对目标运动进行学习,并在隐空间进行节奏化插值,最终合成新的节奏化运动。     

基于组织结构的纺织品各向异性建模方法

纺织面料的物理仿真将虚拟现实与纺织CAD相结合,是虚拟服装与试衣系统研究的一项关键技术,纺织品建模也是纺织品真实感模拟的基础及关键工作.在基于物理的粒子系统的基础上,提出各向异性的建模方式.根据织物面料的不同组织方式进行分析,对其进行经纬方向的分离建模.仿真过程分为初始过程的刚性物体模拟和接近平衡时的柔性物质模拟两大步骤.为了在产生逼真表现效果的同时兼顾计算复杂性,构模时引入多分辨率模型.     

浅海中声源激发的波场成分及特性分析

为更好地认识和利用浅海波导中的声场,该文对浅海中声源激发的波场成分及特性进行了研究.提出了能给出浅海中声场全波解的理论研究方法,给出了声场的复积分表达式,并在复平面上利用围道积分对声场复积分式进行求解,得出了浅海中声源激发出的声场组成部分;同时应用高阶交错网格有限差分法对浅海中声场进行数值模拟,呈现出了不同海水深度、声源频率和声源深度下浅海波导中的波场结构和空间能量分布.研究表明:浅海中声场由离散谱部分和连续谱部分组成;其中离散谱部分包括各阶简正波和Scholte波,连续谱部分包括侧面波;简正波和Scholte波的振幅与水平传播距离的开方成反比,而侧面波的振幅与水平传播距离的平方成反比;海水越浅、声源频率越小、声源深度越大,都会导致海水中的能量越少,越有利于Scholte波的激发,此时声源辐射的能量主要以Scholte波的形式传播出去,能量更多地集中在海底界面处.     

A hybrid method for learning multi-dimensional Bayesian network classifiers based on an optimization model

Bayesian networks, which have a solid mathematical basis as classifiers, take the prior information of samples into consideration. They have gained considerable popularity for solving classification problems. However, many real-world applications can be viewed as classification problems in which instances have to be assigned to a set of different classes at the same time. To address this problem, multi-dimensional Bayesian network classifiers (MBCs), which organize class and feature variables as three subgraphs, have recently been proposed. Because each subgraph has different structural restrictions, three different learning algorithms are needed. In this paper, we present for the first time an MBC learning algorithm based on an optimization model (MBC-OM) that is inspired by the constraint-based Bayesian network structure learning method. MBC-OM uses the chi-squared statistic and mutual information to estimate the dependence coefficients among variables, and these are used to construct an objective function as an overall measure of the dependence for a classifier structure. Therefore, the problem of searching for an optimal classifier becomes one of finding the maximum value of the objective function in feasible fields. We prove the existence and uniqueness of the numerical solution. Moreover, we validate our method on five benchmark data sets. Experimental results are competitive, and outperform state-of-the-art algorithms for multi-dimensional classification.     

Biomechanics-based reaching optimization

This paper presents a Biomechanics-based Lifelike Reaching Controller (BLRC) to generate lifelike reaching motion. The BLRC employs various reaching strategies borrowed from biomechanics to guarantee the naturalness of reaching motion and expands the reachable space to enrich the flexibility of human behavior. We exploit the arm-reachable workspace to guide the motion sampling, and construct different low-dimensional space for each reaching strategy by PCA to reduce the search space, so as to make BLRC fast deal with huge mocap data set. Moreover, we also use the optimization method in these low-dimensional spaces to further speed up the convergence of motion synthesis with the help of the accurate starting point in data space during the search process. We demonstrate the power of the BLRC with more lifelike and complex reaching motion.