针对云计算环境下存在虚拟化资源利用率不高、延迟、性能衰减等问题,提出一种基于粒子群的虚拟资源分配优化方法。根据云计算优势,对虚拟化资源进行描述及状态定义,分析了虚拟环境下的服务质量(quality of services,QoS)组成和计算方法...针对云计算环境下存在虚拟化资源利用率不高、延迟、性能衰减等问题,提出一种基于粒子群的虚拟资源分配优化方法。根据云计算优势,对虚拟化资源进行描述及状态定义,分析了虚拟环境下的服务质量(quality of services,QoS)组成和计算方法;同时通过用户与云提供商收益最大化的平衡关系,建立一种虚拟化操作方法,以确保虚拟化资源被合理分配和回收;采用粒子群算法使用户和虚拟服务提供方都能趋于收益平衡。在此基础上设计的一款云播放器与直接嵌入到Web播放器相比得出,提出的粒子群优化方法有效且将资源利用率至少提高10%,与QoS约束的二元整数规划方法相比,提出的方法操作虚拟资源更有效。展开更多
This paper addresses the challenge of synchronized multiple spacecraft attitude reorientation in presence of pointing and boundary constraints with limited inter-spacecraft communication link.Relative attitude pointin...This paper addresses the challenge of synchronized multiple spacecraft attitude reorientation in presence of pointing and boundary constraints with limited inter-spacecraft communication link.Relative attitude pointing constraint among the fleet of spacecraft has also been modeled and considered during the attitude maneuvers toward the desired states.Formation fling control structure that consists of decentralized path planners based on virtual structure approach joint with discrete time optimal local controller is designed to achieve the mission's goals.Due to digital computing of spacecraft's onboard computer,local optimal controller based on discrete time prediction and correction algorithm has been utilized.The time step of local optimal algorithm execution is designed so that the spacecraft track their desired attitudes with appropriate error bound.The convergence of the proposed architecture and stability of local controller's tracking error within appropriate upper bound are proved.Finally,a numerical simulation of a stereo imaging scenario is presented to verify the performance of the proposed architecture and the effectiveness of the algorithm.展开更多
文摘针对云计算环境下存在虚拟化资源利用率不高、延迟、性能衰减等问题,提出一种基于粒子群的虚拟资源分配优化方法。根据云计算优势,对虚拟化资源进行描述及状态定义,分析了虚拟环境下的服务质量(quality of services,QoS)组成和计算方法;同时通过用户与云提供商收益最大化的平衡关系,建立一种虚拟化操作方法,以确保虚拟化资源被合理分配和回收;采用粒子群算法使用户和虚拟服务提供方都能趋于收益平衡。在此基础上设计的一款云播放器与直接嵌入到Web播放器相比得出,提出的粒子群优化方法有效且将资源利用率至少提高10%,与QoS约束的二元整数规划方法相比,提出的方法操作虚拟资源更有效。
文摘This paper addresses the challenge of synchronized multiple spacecraft attitude reorientation in presence of pointing and boundary constraints with limited inter-spacecraft communication link.Relative attitude pointing constraint among the fleet of spacecraft has also been modeled and considered during the attitude maneuvers toward the desired states.Formation fling control structure that consists of decentralized path planners based on virtual structure approach joint with discrete time optimal local controller is designed to achieve the mission's goals.Due to digital computing of spacecraft's onboard computer,local optimal controller based on discrete time prediction and correction algorithm has been utilized.The time step of local optimal algorithm execution is designed so that the spacecraft track their desired attitudes with appropriate error bound.The convergence of the proposed architecture and stability of local controller's tracking error within appropriate upper bound are proved.Finally,a numerical simulation of a stereo imaging scenario is presented to verify the performance of the proposed architecture and the effectiveness of the algorithm.
