计算机工程与应用 ›› 2016, Vol. 52 ›› Issue (15): 13-18.

• 热点与综述 • 上一篇    下一篇

基于VxWorks下PNN故障诊断的冗余容错系统

李  丹,胡晓光   

  1. 北京航空航天大学 虚拟现实技术与系统国家重点实验室,北京 100191
  • 出版日期:2016-08-01 发布日期:2016-08-12

Redundant fault-tolerant system based on VxWorks and fault diagnosis of PNN

LI Dan, HU Xiaoguang   

  1. State Key Laboratory of Virtual Reality Technology and Systems, Beihang University, Beijing 100191, China
  • Online:2016-08-01 Published:2016-08-12

摘要: 针对航天测控系统的可靠性需求,提出了一种紧凑型PCI总线测控系统的冗余容错设计方案。系统下位机采用了基于VxWorks嵌入式操作系统来保证实时性,并在VxWorks系统中实现了高可用热插拔技术用于提高系统的冗余容错性能。提出了利用基于概率神经网络(PNN)的故障诊断方法对热冗余设备进行在线故障诊断。仿真与实验验证的结果表明,该系统具有良好的冗余容错性能,该设计方法可以有效提升系统的可靠性。

关键词: 冗余, 容错, 故障诊断, 热插拔, VxWorks, 概率神经网络, 紧凑型PCI总线

Abstract: Aiming at the requirements of high reliability for aerospace measurement and control systems, a design of redundancy and fault tolerance based on Compact PCI (CPCI) bus is proposed. The upper computer adopts the embedded system based on VxWorks to ensure the real-time performance. And the high availability hot-swap technology based on VxWorks embedded system is used in the system and an advanced fault diagnosis method based on the Probabilistic Neural Networks(PNN) is proposed for checking the faults of hot-redundancy equipments on line. Simulation and experiment results show that the system can satisfy the requirement of redundancy and fault tolerance. This design method can effectively improve the reliability of the system.

Key words: redundancy, fault-tolerance, fault diagnosis, hot-swap, VxWorks, probabilistic neural networks, compact PCI bus