Research Of Health Management And Multi-Task Scheduling Optimization Under IMA Avionic System

Integrated Modular Avionics(IMA)is the trend of development of avionics system.In order to assure flight security,prognostic and health management(PHM)is an important system based on modern information technology and artificial intelligence.However,PHM is only implemented with respect of health management and emphasize the importance only on fault diagnose and built-in test of a certain device.Also,there is currently no standard specification for PHM system and the research for PHM for aircraft is almost theory-based.Therefore,this paper try to refer to the conception of model-based system engineering,and start from analyzing the system requirement,defining the system function and try to describe the architecture of PHM system based on system modeling language(SysML).Simulation-based experiment shows that PHM with prognosis module can significantly improve the reliability of aircraft device.The second part of this paper aims to optimize multi-task scheduling algorithm in avionics system.For safety concern,both time and space partition are used in IMA to keep real-time tasks in different module independent from other module.However,inappropriate partition allocation scheme may result in huge time redundancy,even cause failure in task scheduling.We try to analyze the way that partition design scheme influence on task scheduling and then propose an optimized scheduling algorithm based on partition adjustment and task re-mapping.Experiment shows that our algorithm can significantly reduce the scheduling failure rate as well as average worst-case average response time of given task set.