Task scheduling and virtual machine allocation policy in cloud computing environment

Cloud computing represents a novel computing model in the contemporary technology world. In a cloud system, the com- puting power of virtual machines （VMs） and network status can greatly affect the completion time of data intensive tasks. How- ever, most of the current resource allocation policies focus only on network conditions and physical hosts. And the computing power of VMs is largely ignored. This paper proposes a comprehensive resource allocation policy which consists of a data intensive task scheduling algorithm that takes account of computing power of VMs and a VM allocation policy that considers bandwidth between storage nodes and hosts. The VM allocation policy includes VM placement and VM migration algorithms. Related simulations show that the proposed algorithms can greatly reduce the task comple- tion time and keep good load balance of physical hosts at the same time.

Multilevel manufacturing system of virtual enterprise based on manufacturing grid and strategies for member enterprise selection and task assignment

In order to improve efficiency of virtual enterprise, a manufacturing grid and multilevel manufacturing system of virtual enterprise is built up. When selecting member enterprises and task assignment based on the manufacturing grid, key activities are assigned to the suitable critical member enterprises by task decomposition, enterprise node searching and characteristic matching of manufacturing resources according to the characteristic matching strategy. By task merger, some ordinary activities are merged with corresponding key activities and assigned to corresponding critical member enterprises. However, the other ordinary activities are assigned to the related ordinary member enterprises with enterprise node searching and characteristic matching of manufacturing resources. Finally, an example of developing the artificial hip joint in the virtual enterprise is used to demonstrate that efficiency of the virtual enterprise is improved by using the manufacturing grid and the proposed strategies for member enterprise selection and task assignment.

An fNIRS-Based Study on Prefrontal Cortex Activity during a Virtual Shopping Test with Different Task Difficulties in Brain-Damaged Patients

We developed a Virtual Shopping Test with three different task levels for assessment of daily cognitive function using virtual reality technology. The objective of present study was to investigate the difference on task performance, brain activation and subjective assessment in relation to the difficulty levels of the tasks. Subjects were asked to buy specific 2 items in Task 1, 4 items in Task 2, and 6 items in Task 3 at a virtual mall. The tasks and questionnaires were conducted on 10 convalescent brain-damaged patients and 6 healthy young adults. Hemodynamic changes in the prefrontal cortex (PFC) during activation due to the tasks were examined using functional near-infrared spectroscopy. As the result, the mean total time was significantly longer for the patients than for healthy subjects. PFC showed a greater response for related Task 2 than Task 1 in shopping and moving phase in patient group. The patients evaluated Tasks 1 and 2 are more difficult and bring more psychological load than healthy adults subjectively. That is, although the healthy adults did not show large difference in their task performances as well as PFC responses, they can evaluate the differences between three task levels, subjectively, while which could not be for the patients means that patients could not distinguish the difference of the tasks, subjectively. The results suggest that 4-item shopping task might be enough difficulty level that causes brain activation for the brain-damaged patients.

基于虚拟空间三维体素—语义地图的机械臂任务规划

任务规划算法是机械臂在未知环境下具备抓取能力的基础,为此提出一种基于三维体素—语义地图的机械臂虚拟空间任务规划方法。在获取完整点云场景的基础上,利用基于掩码区域的卷积神经网络实现目标检测与实例分割,构建由三维点云、语义信息、三维计算机辅助设计模型组成的体素-语义混合地图;优化A*算法完成机械臂末端执行器最优路径规划,利用贝塞尔曲线完成轨迹优化;探究抓取不同几何体时的机械臂末端位姿,分解抓取、放置等复杂任务,形成基于语义驱动的空间任务规划。在虚拟空间通过仿真试验,表明了所提算法的有效性和快速性,同时该算法可支撑机械臂执行各种复杂的抓取任务,并提升机械臂的智能水平。

旧城社区景观更新设计虚拟仿真实验的建设与实践

针对景观更新设计实验教学中存在的空间跨度大、项目周期长、投资高昂且过程不可逆等难题,设计开发了旧城社区景观更新设计虚拟仿真实验。实验由知识体系、任务体系与评价体系构成,采取“项目引领、任务驱动”的总设计思路,通过对景观更新典型案例的拆解与模块化建设,实现了知识点与设计全过程高仿真可视化呈现;同时构建了多元量化评价体系,对设计方案的经济性、生态性、艺术性等进行实时在线评估,不断强化行业规范和标准意识,树立系统性与融合性的设计理念,引导学生统筹协调各设计要素,反复优化直至形成最优方案,有效提升了学生的总体设计能力和解决复杂工程问题的能力。

基于HTN的巡视器动态规划方法

为提升深空探测任务中巡视器动态规划处理效率,从航天工程实践经验出发,结合分层任务网络(hierarchical task network,HTN)理论研究成果,提出巡视器领域知识建模方法以及动态规划流程。针对经典动作规划系统不能处理时态约束关系的问题,在建模中引入受限状态事件的表示方法。针对动态规划中的时态约束关系,拓展Allen区间代数的定义,提出事件、全局时间点和虚拟事件的时序逻辑判断方法。针对动态规划中存在的状态空间一致性问题,提出虚拟指令继承原有状态空间的方法。研究结果完善了HTN在航天领域的知识建模方法,解决了经典动作规划缺少时间信息约束的缺陷。所提出的虚拟事件、虚拟指令解决动态规划状态判定方法,可有效提高深空探测任务中巡视器遥操作的工作效能和地面中心进行事件处理的灵活性。

混合现实中基于GPU虚拟化的AI计算优化

研究探讨混合现实(MR)应用中,通过GPU虚拟化优化AI计算,聚焦于多任务调度与资源共享。研究提出了一个模型,其包含一种根据任务优先级、资源需求和等待时间,动态为正在执行的任务分配GPU资源的机制。同时,模型采用优化的多任务调度算法,以提高调度效率。实验结果表明,尽管在单任务性能测试中模型的执行时间、GPU利用率和内存使用方面略逊于物理GPU,但在多任务并发和资源共享方面,研究提出的模型展现了显著优势。未来研究将探索设计更公平高效的资源共享策略,以及进一步优化多任务调度算法。

基于虚实结合的任务信息引接系统设计与实现

为满足训练的信息化需求,设计了灵活通用的数据协议和虚实结合的信息引接系统。基于中间件的分层设计思路,实现了外部信息接收、对内信息转发、通信协议解析与转换、信息对接联调、数据文件记录、配置管理以及信息流量统计等功能;采用了MVC软件架构模式,选用Visual Studio 2015作为开发工具,以Oracle数据库作为底层支撑,构建了实装信息及虚拟目标信息接收模块、建立数据映射关系模块、读取数据库方案模块、基础通信接收内部航迹模块、数据包发送模块。实际应用效果表明,该系统能够屏蔽外部装备系统、仿真系统和网络协议之间的差异,提供互联互通,适应虚实结合的训练任务,提高保障能力。该系统具有良好的适应性和可扩展性,可以满足信息化条件下系统的应用和发展需要。

基于偏好和虚拟适应度的两阶段依赖任务卸载算法

针对云边端协同环境中依赖任务卸载时效率低以及任务卸载失败的问题,提出一种基于偏好和虚拟适应度的两阶段依赖任务卸载算法.第一阶段,根据提出的二维卸载偏好因子对依赖任务的部分子任务进行直接卸载决策,从而有效缩小遗传算法初始种群的规模.第二阶段,提出基于虚拟适应度的启发式交叉方法,并对基于参考点的快速非支配排序遗传算法(non-dominated sorting genetic algorithmⅢ, NSGA-Ⅲ)的交叉算子进行改进,保留了种群多样性并提升了算法收敛速度,最后使用改进的算法对所有依赖任务的子任务进行最优卸载决策集的搜索.实验结果表明,与其他算法相比,该算法在任务完成时间、任务能耗和边缘云集群成本方面平均优化了10.2%~18.3%,并且将任务失败率平均降低了10.7%~25.6%.

一种用于星载虚拟化平台的任务容器调度算法

星载虚拟化平台借助容器等轻量级虚拟化技术,将计算任务封装到容器中形成任务容器,从而实现资源的高效利用。然而,该平台的任务容器调度问题是一个亟需解决的难题。针对这一问题建立了一个基于非阻塞通信模式的可分容器任务多趟调度模型。在该基础上,提出了一种新的调度算法,旨在确定最佳的处理机调度顺序和调度趟数。该算法结合可分任务容器和多趟调度的概念,通过将任务分解为可执行的子任务,并在多个调度阶段中进行任务分配和处理机调度,从而优化调度顺序,提高整体处理效率。该算法是一种改进的遗传算法,在传统遗传算法的基础上添加了子种群隔离的优化策略,其核心思想是将种群划分策略引入算法过程,从而改善遗传算法的性能和效果。通过实验验证了该算法的有效性和收敛性,结果表明,该算法缩短了任务完成时间。

交叉融合的新工科机器人焊接实训课程体系规划与实践研究

针对新工科建设培养交叉融合人才的目标,提出了机器人焊接实训课程的设计与规划。合理规划设计教学流程,将机器人焊接工作站操作与虚拟仿真编程实训相结合,充分利用虚拟仿真软件提高实训教学效率。分析机器人焊接实训设备适配的实训交叉学科知识内容,将焊枪动作轨迹编程与机器人运动学、夹具动作与自动控制、金属材料弧焊与焊机控制等交叉技术之间相互链接,建设新工科交叉融合的机器人焊接实训课程体系。对接企业技术岗位需求,规划能力提升典型训练任务,提高学生应用交叉技术解决实际问题的应用能力,培养满足产业需求的新工科应用型人才。

虚拟仪器的电机测试系统设计研究

由于目前的微型计算机通用试验系统大都是使用Windows的多媒体软件,其计时精度和实时性都很差。本文提出利用虚拟仪器技术对电机测试试验系统进行仿真,并把软件工程学的概念引入到该试验系统的设计中,实现了各部分的功能相对的独立性,从而保证了该试验系统的可靠性。在此基础上,借助LabVIEW自身的特性,实现了基于操作系统的较为准确的计时工作,并具备了很强的移植能力。实践证明,此系统性能可靠、操作简便、实用方便,同时也减少了试验费用,能够很好地适应现场试验的需要。

基于多任务判别器与注意力机制的虚拟试衣研究

为了解决具有错位和遮挡处理条件的高分辨率虚拟试戴(HR-VITON)在处理复杂纹理表现和服装特征交互方面的局限性问题,在基于具有错位和遮挡处理条件的高分辨率虚拟试衣方法的基础上,提出了一种结合多任务判别器与注意力机制的虚拟试衣方法。首先,通过在条件构造器中加入高效通道注意力机制,有效地增强了特征融合;其次,在图像生成网络中采用多任务判别器,以增强对服装渲染的全局和局部尺度评估。通过不断调整网络的学习参数,最终将模型放在数据集VITON-HD Dataset上进行虚拟试衣实验。实验结果表明,与原方法相比,该方法的图像感知相似度(LPIPS)提升了6%、分布距离指标(FID)提升了4.8%,虚拟试衣效果更好。

基于虚拟传感器的坝区多输出自由场地震时程长序列预测模型研究

坝区自由场地震时程的多维长时序预测对于震害快速分析具有重要意义。虚拟传感器是地震物理传感器的补充感知手段,可实现地震时程的预测,然而现有虚拟传感器难以对多个信号做长时序预测,导致大坝震害分析较为滞后。针对上述问题,提出基于TFA-Seq2Seq虚拟传感器的坝区多输出自由场地震时程长序列预测模型。其中,基于多任务学习将Seq2Seq的虚拟传感器改进为“Encoder-3 Decoder”结构,以建立多个坝体物理传感器信号与自由场三个方向长时序地震时程的映射关系,并添加注意力机制捕获多个输入信号的时序依赖关系,以解决同步多输出预测问题及提升预测精度。进一步,引入可逆的时频变换层和其逆变换层改进编码器和解码器,以缩短地震信号的时域长度,提取频域特征,并提出对应的随机强制学习的模型训练策略,从而克服了现有虚拟传感器难以对长序列进行有效预测的缺陷。案例分析表明,该方法实现了坝区自由场三个方向地震信号的超前10 s虚拟感知,且相较于未添加注意力机制和单输出的模型,预测精度分别提高了6.88%和3.32%,研究为震时地震信息的超前感知提供了新思路和新途径。

面向航空发动机的虚拟多人拆装任务的驱动方法

为满足虚拟多人协同拆装系统不同拆装任务的快速开发需求,提出一种可配置任务序列的数据驱动方法。以某型航空发动机样机为研究对象,研究了多人协同拆装的任务要素,对任务进行基于JS对象图谱(JavaScript Object notation,JSON)的参数化设计与表达,以JSON参数化文件驱动任务序列,定义每一步任务的交互操作。实践表明:该方法运用于多人协同拆装系统中,使系统具备可配置性,能根据自定义任务序列驱动任务流程;在多端大任务量的情况下,有效提升部署效率。

Integrated Virtual Assembly Process Planning System

Assembly process planning（APP） for complicated products is a time-consuming and difficult work with conventional method. Virtual assembly process planning（VAPP） provides engineers a new and efficiency way. Previous studies in VAPP are almost isolated and dispersive, and have not established a whole understanding and discussed key realization techniques of VAPP from a systemic and integrated view. The integrated virtual assembly process planning（IVAPP） system is a new virtual reality based engineering application, which offers engineers an efficient, intuitive, immersive and integrated method for assembly process planning in a virtual environment. Based on analysis the information integration requirement of VAPP, the architecture of IVAPP is proposed. Through the integrated structure, IVAPP system can realize information integration and workflow controlling. In order to mode/the assembly process in IVAPP, a hierarchical assembly task list（HATL） is presented, in which different assembly tasks for assembling different components are organized into a hierarchical list. A process-oriented automatic geometrical constraint recognition algorithm（AGCR） is proposed, so that geometrical constraints between components can be automatically recognized during the process of interactive assembling. At the same time, a progressive hierarchical reasoning（PHR） model is discussed. AGCR and PHR will greatly reduce the interactive workload. A discrete control node model（DCNM） for cable harness assembly planning in IVAPP is detailed. DCNM converts a cable harness into continuous flexed line segments connected by a series of section center points, and designs can realize cable harness planning through controlling those control nodes. Mechanical assemblies （such as transmission case and engine of automobile） are used to illustrate the feasibility of the proposed method and algorithms. The application of IVAPP system reveals advantages over the traditional assembly process planning method in shortening the time-consumed in assembly planning and in minimizing the handling difficulty, excessive reorientation and dissimilarity of assembly operations.

Realization Techniques of Virtual Assembly Process Planning System

The key realization techniques of virtual assembly process planning （VAPP） system are analyzed, including virtual assembly model, real-time collision detection, automatic constraint recognition algorithm, cable harness assembly process planning and visual assembly process plan at the workshop. A virtual assembly model based on hierarchical assembly task list （HATL） is put forward, in which assembly tasks are defined to express component assembling operations and are sequentially and hierarchically organized according to different subassemblies, which can perfectly model the construction process of product, And a multi-layer automatic geometry constraint recognition algorithm of how to identify assmbly constraint relations in the virtual environment is proposed, then a four-layer collision detection algorithm is discussed. A VAPP system is built and some simple mechanical assemblies are used to illustrate the feasibility of the proposed method and algorithms.

Creation and Evaluation of a Multi-sensory Virtual Assembly Environment

The multi-modal information presentation, integrated into the virtual environment （VE）, has potential for stimulating different senses, improving the user＇s impression of immersion, and increasing the amount of information that is accepted and processed by the user＇s perception system. The increase of the useful feedback information may reduce the user＇s cognitive load, thus enhancing the user＇s efficiency and performance while interacting with VEs. This paper presents our creation of a multi-sensory virtual assembly environment （VAE） and the evaluation of the effects of multi-sensory feedback on the usability. The VAE brings together complex technologies such as constraint-based assembly simulation, optical motion tracking technology, and real-time 3D sound generation technology around a virtual reality workbench and a common software platform. The usability evaluation is in terms of its three attributes： efficiency of use, user satisfaction, and reliability. These are addressed by using task completion times （TCTs）, questionnaires, and human performance error rates （HPERs）, respectively. Two assembly tasks have been used to perform the experiments, using sixteen participants. The outcomes showed that the multi-sensory feedback could improve the usability. They also indicated that the integrated feedback offered better usability than either feedback used in isolation. Most participants preferred the integrated feedback to either feedback （visual or auditory） or no feedback. The participants＇ comments demonstrated that nonrealistic or inappropriate feedback had negative effects on the usability, and easily made them feel frustrated. The possible reasons behind the outcomes are also analysed by using a unifying human computer interaction framework. The implications, concluded from the outcomes of this work, can serve as useful guidelines for improving VE system design and implementation.

面向云雾计算的实时任务调度算法的设计及分析

相比于云计算,雾计算通过将计算、存储和网络服务靠近物联网设备,满足了时延和响应时间要求。为此,提出基于模糊逻辑的实时任务调度(fuzzy logic-based real-time task scheduling,FLTS)算法。先利用模糊逻辑算法合理地给云层和雾层分配任务,当雾层执行任务时,就依据任务对计算、存储和带宽要求以及任务的截止日期、数据尺寸,选择最适宜的虚拟机处理任务。仿真结果表明,相比于同类算法,提出的FLTS算法提升了执行任务的成功率、完成时间以及平均周转时间等性能。