基于蚁群算法的分布式CPS系统任务调度设计  

Design of Distributed CPS System Task Scheduling Based on Ant Colony Algorithm

作　　者：韩杰[1] 马斌[1] 黄宽[1] 王长涛[1] 魏威[1] 

机构地区：[1]沈阳建筑大学信息与控制工程学院,辽宁沈阳110168

出　　处：《测控技术》2015年第2期121-124,共4页Measurement & Control Technology

基　　金：住建部2012年科学技术项目(2012-K8-33)

摘　　要：针对CPS系统中资源和任务的复杂性,同时为了保证CPS系统资源管理和任务调度的快速性、准确性和有效性,构建了分布式CPS系统结构模型,并设计和实现了基于蚁群算法的资源分配策略,在算法中设计了局部和全局信息素更新规则,将资源管理、任务调度和执行相结合,同时兼顾网络的负载平衡问题和服务质量问题。利用Matlab对基于蚁群算法的分布式CPS系统任务调度策略进行了仿真实验,实验结果证明该策略缩短了整体任务调度的总完成时间,并随着任务数的增加,任务时间缩短的效果更加明显。For the complexity of the resources and tasks in CPS,meanwhile,in order to ensure the rapidity,accuracy and validity of resource management and task scheduling in CPS,a distributed architecture model of CPS is built,and resource allocation strategy based on ant colony algorithm is designed and implemented.In the algorithm,the update rules of local and global pheromone are designed,combining with the resource management,task scheduling and execution,and considering both of network load balancing and service quality.Using Matlab to do the simulating experiment about the design of distributed CPS＇s task scheduling based on ant colony algorithm,the results show that the strategy shortens the total completion time of the overall task scheduling,and with the increasing in the number of tasks,the completion time is shortened better significandy.

关 键 词：CPS 分布式计算 蚁群算法 任务调度 智能建筑 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]