摘要
以往的异构多核嵌入式系统实时调度方法由于仅设置了系统实时调度模型的单项参数,导致系统调度时间过长。本文设计了面向任务协同的异构多核嵌入式系统实时调度方法。面向任务协同构建嵌入式系统实时调度模型,计算系统任务节点的传输情况,构建系统实时调度的数学模型,对构建的实时调度模型中的任务数据信息进行汇集,并进行自适应分析处理,以此为基础,设置实时调度模型的时间均衡控制参数和安全性系数,从而实现嵌入式系统的实时调度。通过上述设计,完成对异构多核嵌入式系统实时调度方法的设计。在仿真实验中,与以往的异构多核嵌入式系统实时调度方法相比,本文设计的面向任务协同的异构多核嵌入式系统实时调度方法的调度时间最长仅为5 s,调度时间更短。
Previous real-time scheduling methods for heterogeneous multi-core embedded systems resulted in excessively long system scheduling time due to setting single parameters of the real-time scheduling model.Therefore,a real-time scheduling method for heterogeneous multi-core embedded systems oriented towards task collaboration was designed.Firstly,a real-time scheduling model for embedded systems is constructed for task collaboration.Then,the transmission status of system task nodes is calculated,and a mathematical model for real-time scheduling is constructed.The task data information in the constructed real-time scheduling model is collected and adaptively analyzed and processed.Based on this,the time balance control parameters and security coefficients of the real-time scheduling model are set to achieve real-time scheduling of embedded systems.Through the above design,the design of a real-time scheduling method for heterogeneous multi-core embedded systems is completed.In simulation experiments,compared with previous real-time scheduling methods for heterogeneous multi-core embedded systems,the designed real-time scheduling method for task collaboration oriented heterogeneous multi-core embedded systems has a maximum scheduling time of only 5 seconds and a shorter scheduling time.
作者
程玮
杨智玲
CHENG Wei;YANG Zhiling(Xiamen Ocean Vocational College,361100 Xiamen,China;Xiamen Key Laboratory of Intelligent Fishery,361100 Xiamen,China)
出处
《长春师范大学学报》
2024年第2期43-49,共7页
Journal of Changchun Normal University
基金
中国高校产学研创新基金新一代信息技术创新项目“无人机技术在水产养殖中的应用研究”(2021ITA10012)
福建省海洋经济发展专项资金项目“海水养殖尾水高效治理装备开发与应用”(FJHJF-L-2021-12)
厦门海洋职业技术学院高层次人才研究项目“高超音速飞行器非线性控制方法研究”(KYG202101)。
关键词
任务协同
异构多核
嵌入式系统
实时调度
方法设计
task collaboration
heterogeneous multi-core
embedded system
real time scheduling
method design
