基于自适应t分布的改进粒子群实时任务调度算法  

Real-time Task Scheduling Algorithm of the Improved Particle Swarm Based on the Self-adaptive t-Distribution

作　　者：柳子来 王健敏[2] LIU Zi-lai;WANG Jian-min(Faculty of Information Engineering and Automation,Yunnan Key Laboratory of Artificial Intelligence,Kunming University of Science and Technology;Yunnan Rural Science and Technology Service Center)

机构地区：[1]昆明理工大学信息工程与自动化学院云南省人工智能重点实验室 [2]云南省农村科技服务中心

出　　处：《化工自动化及仪表》2020年第5期393-397,424,共6页Control and Instruments in Chemical Industry

摘　　要：针对信息物理融合系统(CPS)中实时任务调度存在效率低和无法满足用户多种服务质量(QoS)需求的问题,提出一种基于自适应t分布的改进粒子群实时任务调度算法(t-PSO)。首先,该算法在传统粒子群算法(PSO)的基础上,引入了自适应t分布的变异机制,达到提高收敛速度和避免算法陷入局部最优的目的。其次,以任务完成时间、任务总成本和服务质量为衡量标准设置适应度函数来完成任务调度。最后,与粒子群算法、柯西变异粒子群算法(Cauchy-PSO)进行任务调度仿真对比实验。结果表明:相同的实验条件下,t-PSO算法具有更好的整体性能,在任务完成时间、任务总成本和服务质量上的表现都明显优于其他两种算法。In cyber-physical system(CPS),the real-time task scheduling has low efficiency and fails to meet user’s multiple quality of service(QoS),aiming at those problems,an improved particle swarm real-time task scheduling algorithm(t-PSO)based on self-adaptive t-distribution was proposed.Firstly,basing on traditional PSO algorithm,the t-PSO proposed has the mutation mechanism of self-adaptive t-distribution adopted to improve convergence speed and avoid algorithm to fall into local optimization;and then,it has task completion time,task’s total cost and service quality taken as the reference to set a fitness function for the task scheduling;finally,based on a task scheduling simulation experiment on the cyber-physical system(CPS),it has t-PSO algorithm compared with basic PSO algorithm and Cauchy mutation particle swarm optimization(Cauchy-PSO)algorithm to show that,under the same experimental conditions,the t-PSO algorithm proposed has better overall performance,and it outperforms other two algorithms in task completion time,task total cost and service quality.

关 键 词：信息物理融合系统 自适应t分布 粒子群优化 实时任务调度 服务质量 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]