摘要： 在某些软实时系统中，为了降低系统成本，允许系统过载，并且允许过载时少量任务能错失时限；对于这样的系统，常常能设法获得其任务的执行时间所服从的概率分布。针对这样的系统，提出了基于概率的可调度分析方法，对系统的性能作出概率上的估计，在概率意义上预测系统的可行性；并提出以数据挖掘技术估算任务执行时间的分类方法，这对预测系统的性能和提高系统利用率有重要的意义。

关键词: 实时系统；实时任务调度；概率密度；数据挖掘

Abstract: In some soft real-time systems, tasks can miss their deadlines especially when the systems are overloaded, and the probability of execution time of tasks can be known. This paper provides a method of schedulability analysis based on probability for such systems, and describes how to predict the performance and estimates the execution time of tasks by data-mining, which will be contributive to the improvement of the utility of the systems.

Key words: Real-time systems; Real-time task scheduling; Probability; Data mining