华南理工大学学报(自然科学版) ›› 2010, Vol. 38 ›› Issue (9): 30-34.doi: 10.3969/j.issn.1000-565X.2010.09.006

• 电子、通信与自动控制 • 上一篇    下一篇

基于主动请求与动态分配负载的CAN容错算法

曹小华 周勇   

  1. 武汉理工大学 物流工程学院, 湖北 武汉 430063
  • 收稿日期:2009-12-21 修回日期:2010-05-20 出版日期:2010-09-25 发布日期:2010-09-25
  • 通信作者: 曹小华(1973-),男,博士,副教授,主要从事智能控制与通信技术研究. E-mail:Tomm_cao@163.com
  • 作者简介:曹小华(1973-),男,博士,副教授,主要从事智能控制与通信技术研究.
  • 基金资助:

    国家自然科学基金资助项目(50975216); 武汉理工大学中央高校基本科研业务费专项资金资助项目(2010-1a-059)

Fault-Tolerant Algorithm Based on Active Request and Dynamic Load Distribution for CAN System

Cao Xiao-hua  Zhou Yong   

  1. School of Logistics Engineering,Wuhan University of Technology,Wuhan 430063,Hubei,China
  • Received:2009-12-21 Revised:2010-05-20 Online:2010-09-25 Published:2010-09-25
  • Contact: 曹小华(1973-),男,博士,副教授,主要从事智能控制与通信技术研究. E-mail:Tomm_cao@163.com
  • About author:曹小华(1973-),男,博士,副教授,主要从事智能控制与通信技术研究.
  • Supported by:

    国家自然科学基金资助项目(50975216); 武汉理工大学中央高校基本科研业务费专项资金资助项目(2010-1a-059)

摘要: 控制器局部网(CAN)总线应用于负载较重的分布式系统时,系统的可靠性和实时性难以保证,容错算法是提高分布式系统容错性能的重要手段。文中通过对多种容错算法的分析比较,在冗余服务中引入主动请求与动态分配负载的思想,提出了一种有效的CAN冗余容错算法.CAN节点可根据自身的负载情况决定是否作为冗余服务器主动请求任务,管理系统综合考虑负载和节点主机的当前状况,动态分配负载给各冗余服务器,同时系统还可动态改变冗余服务器的个数.最后,构建了该算法的数学模型,与RAWA算法进行了对比,并通过实例证明该算法性能较好,可用于高可靠、高实时性的控制系统中.

关键词: 容错, 控制器局部网, 动态负载, 主动请求

Abstract:

When the controller area network(CAN) bus is applied to a distributed overloading system,the reliability and real-time property of the system can not be guaranteed.Fault-tolerant algorithms are the important means to improve the system performances.In this paper,several fault-tolerant algorithms are analyzed and compared,and an approach of introducing active request and dynamic load distribution in the redundant services is proposed,based on which an effective fault-tolerant algorithm for CAN systems is put forward.In this algorithm,CAN nodes decide whether to actively request the task to be a redundant server according to their load status,and the management system assigns the load to each redundant server according to the load and the current status of node hosts and changes the number of redundant servers dynamically.A mathematical model of the proposed algorithm is also established,and some illustrations are finally presented to compare the algorithm with RAWA fault-tolerant algorithms.The results show that the proposed algorithm with high performance is applicable to the control system with high reliability and good real-time property.

Key words: fault tolerance, controller area network, dynamic load, active request