虚拟计算环境下基于聚类的资源匹配优化模型

虚拟计算环境中任务具有数量庞大、需求模糊、种类多样等特征,使得资源匹配面临巨大挑战.依据虚拟计算实验床平台公布数据,提出了一种融合虚拟资源与任务聚类的资源匹配优化模型.该模型通过分析任务需求、消耗等特征,基于改进二分K均值进行任务聚类,并结合虚拟资源类型生成优化的资源匹配列表.经实验分析验证,该模型有效缩小资源匹配范围,提高任务运行成功率,为精准匹配提供基础.     

计算机视觉原理分析及其应用

介绍了计算机视觉的原理,分析解释了实现原理的步骤以及发展现状,并且举例说明了计算机视觉的应用,解决了在人脸检测和识别、机器人目标定位、导航等领域的应用问题,并简单论述了计算机视觉技术的发展方向.     

一种基于深度信息的人头检测方法

针对目前人头检测方法对光线变化敏感和易受阴影干扰的问题,提出了一种基于深度图像的人头检测方法.首先通过运动目标检测,得到运动人员所在区域;然后对该区域使用改进的立体匹配算法,该匹配算法对传统的WTA匹配算法进行改进,只对强纹理点进行匹配,对弱纹理点只进行视差验证,并根据三角投影原理计算出深度图.由于深度图中人员与周围场景的深度分布不同,根据深度分布将人头区域提取出来,得到候选区域,最后将候选区域经过形态学运算并根据区域轮廓的特征来判断是否为人头.实验结果表明:该方法在不同光线环境条件下的检测正确率为94%以上,误检测率仅为5.77%,检测精度高,对光线和阴影的抗干扰性良好,能够很好地适应复杂环境.     

Vertical load capacities of roof truss cross members

Trusses used for roof support in coal mines are constructed of two grouted bolts installed at opposing forty-five degree angles into the roof and a cross member that ties the angled bolts together. The load on the cross member is vertical, which is transverse to the longitudinal axis, and therefore the cross member is loaded in the weakest direction. Laboratory tests were conducted to determine the vertical load capacity and deflection of three different types of cross members. Single-point load tests, with the load applied in the center of the specimen and double-point load tests, with a span of 2.4 m, were conducted. For the single-point load configuration, the yield of the 25 mm solid bar cross member was nominally 98 kN of vertical load, achieved at 42 cm of deflection. For cable cross members, yield was not achieved even after 45 cm of deflection. Peak vertical loads were about 89 kN for 17 mm cables and67 kN for the 15 mm cables. For the double-point load configurations, the 25 mm solid bar cross members yielded at 150 kN of vertical load and 25 cm of deflection. At 25 cm of deflection individual cable strands started breaking at 133 and 111 kN of vertical load for the 17 and 15 mm cable cross members respectively.     

Adaptive key SURF feature extraction and application in unmanned vehicle dynamic object recognition

A new method based on adaptive Hessian matrix threshold of finding key SRUF( speeded up robust features) features is proposed and is applied to an unmanned vehicle for its dynamic object recognition and guided navigation. First,the object recognition algorithm based on SURF feature matching for unmanned vehicle guided navigation is introduced. Then,the standard local invariant feature extraction algorithm SRUF is analyzed,the Hessian Metrix is especially discussed,and a method of adaptive Hessian threshold is proposed which is based on correct matching point pairs threshold feedback under a close loop frame. At last,different dynamic object recognition experiments under different weather light conditions are discussed. The experimental result shows that the key SURF feature abstract algorithm and the dynamic object recognition method can be used for unmanned vehicle systems.     

混沌理论在电力系统负荷预测中应用综述

电力系统负荷的准确预测,对电力系统安全、高效运行有重要作用。负荷存在混沌特性,而基于混沌理论的预测较好地刻画了负荷物理属性,预测精度高。并主要介绍了混沌理论在负荷预测中的应用。首先,概述了基于混沌理论预测的机理,介绍了几种混沌特性判别的方法。然后,介绍了几种预测模型。最后,对现有预测模型总结和展望。     

混合动力车制动工况分析与储能装置参数匹配

为提高混合动力系统整体性能,实现高效能量回收,分析某重型越野车辆驾驶循环工况中制动过程的功率与能量分布,从制动能量回收率与电机参数出发讨论对储能装置的性能要求. 提出电池组-超级电容复合储能装置的参数匹配方法,针对21 t级试验样车混合动力系统进行实例计算,论证锂离子电池组与超级电容组成的复合储能装置的性能. 实例计算与道路试验结果表明:匹配的复合储能装置符合车辆整体性能与制动能量回收的要求,体积、重量满足总体设计约束;匹配超级电容后,储能装置的瞬时功率能力大幅提升,可显著提高车辆的制动能力和制动能量回收率.     

面向大尺度三维重建的多核并行SIFT算法

针对大尺度三维重建中图像数据集较大、其特征提取和匹配时间消耗过长等问题,提出了一种针对有序图像数据集的滑动窗口匹配算法.该算法采用并行计算实现了对SIFT特征提取和匹配算法的加速,并在图形处理器上得以实现,减少了匹配计算量.实验结果表明,算法大大提高了特征提取和匹配的运算速度,保证了特征的稳定性和匹配精度,有效地缩短了大尺度三维重建的运算时间,确保了重建精度.     

基于斜投影算子的压缩采样匹配追踪算法

斜投影匹配追踪算法较正交匹配追踪算法针对非正交字典的重构能力有了大幅提高,但受噪声影响大。该文通过投影矩阵迭代公式导出斜投影匹配追踪算法,并引入回溯的思想,提出基于斜投影算子的压缩采样匹配追踪算法,有效降低了噪声的影响和非正交投影产生的误差,提高了重构精度。仿真结果表明,算法的重构效果优于其它同类算法。     

混合动力客车动力系统参数匹配与仿真优化

针对目前混合动力城市客车所具有的优势,本文以某款城市客车为例,在3种混合动力驱动形式的基础上,确定了本车串联混合动力的驱动形式;并对动力系统各部件参数进行了匹配设计,同时利用ADVISOR软件及其自动尺寸设计功能进行优化仿真。仿真结果表明,优化后的参数,虽电池组数目相差不大,但发动机、发电机和电动机的功率均有所减小,从而使3个部件的尺寸、质量和体积都有所减小,因此,整车的总质量也将随之减小,整车的经济性比参数优化前好。该研究满足了混合动力城市客车对动力性能的要求,达到了参数优化的目的。该研究为发展混合动力客车提供了理论依据。