基于马尔可夫电池模型的节能调度算法研究  

A Power-aware Scheduling Algorithm Based on Markov Battery Model

作　　者：李永亭[1] 齐咏生[1] 

机构地区：[1]内蒙古工业大学信息工程学院,内蒙古呼和浩特010051

出　　处：《计算机仿真》2009年第10期283-288,共6页Computer Simulation

基　　金：内蒙古自然科学基金(200711020810)

摘　　要：对于电池供电的便携式系统,电池的节能调度是提高系统性能的重要因素,而建立精确实用的电池模型是研究电池节能调度算法的必要前提。提出了一种电池建模方法,即在每个单位时间内模拟电池的基本行为,通过行为的叠加实现对电池整个放电过程的跟踪。仿真和实验结果均证明了建模方法的正确性,同时亦表明方法具有建模简单、模型精度较高等优点。针对现有的电池节能算法往往存在增加系统复杂度、影响系统实时性等缺点,文中提出两种改进的电池节能调度算法,并在电池模型上进行了仿真验证。结果显示,改进后的算法更易于实现,且在满足系统实时性能前提下,达到较好的节能效果。Power - aware scheduling of battery is a very important factor to improve the performance of the portable device powered by battery. It is necessary to build a precise and useful battery model for studying Power - aware scheduling algorithm of battery. So this paper puts forward a modeling method which simulates the basic acts of bat- tery in each unit time to trace the discharging process of battery. The results of simulation and experiment demon- strate the accuracy and validity of this approach. And the modeling method is simple and accurate. In view of the shortcomings of current Power - aware scheduling algorithm in increasing the complexity of system and influencing the real - time property of the system, two scheduling algorithms of battery are presented and simulation experiments are done on the basis of the model. The results show that the two scheduling algorithms achieve better power saving effect without decreasing the system real - time property.

关 键 词：电池模型 节能调度算法 离散马尔可夫过程 

分 类 号：TP15[自动化与计算机技术—控制理论与控制工程]