Model-based partition scheduling of integrated modular avionics systems using genetic algorithm

基于模型的集成模块化航空电子系统分区调度及遗传算法

阅读:1

Abstract

This paper proposes a model-based method for partition optimization in Integrated Modular Avionics (IMA) systems that comply with the ARINC 653 standard. It is a complex combinatorial optimization problem to schedule a variety of real-time partitions to achieve maximum processor utilization in an IMA system. Our approach effectively utilizes timed automata in UPPAAL to address the modeling challenges posed by the ARINC 653 features, including two-level scheduling, multi-core processors, and both strictly and loosely periodic partitions. A parallel genetic algorithm is employed to explore the scheduling parameter space of the IMA model, accelerating the computationally expensive model-based analysis through parallelization and achieving an optimal solution more efficiently than traditional analytical methods. Furthermore, we propose a compositional framework that allows for the independent verification of each partition while aggregating local results to derive global schedulability properties, thereby mitigating the state space explosion commonly encountered in model checking. Experimental results demonstrate that our method achieves lower processor occupancy and improved optimization performance compared to traditional techniques, thereby enhancing both the practical applicability of partition scheduling in IMA systems and the overall efficiency of real-time system design.

特别声明

1、本页面内容包含部分的内容是基于公开信息的合理引用;引用内容仅为补充信息,不代表本站立场。

2、若认为本页面引用内容涉及侵权,请及时与本站联系,我们将第一时间处理。

3、其他媒体/个人如需使用本页面原创内容,需注明“来源:[生知库]”并获得授权;使用引用内容的,需自行联系原作者获得许可。

4、投稿及合作请联系:info@biocloudy.com。