基于可不断电维修的双机容错系统

对单个单片机控制系统可靠性不高的问题,提出了一种由双单片机构成容错系统的实现方法.在该系统中,用串行通讯技术,通过采集对方I/O输出,与自身的I/O输出值作比较,实现双单片机的故障互检;利用地址译码技术,设计仲裁切换模块,完成故障隔离;故障机修复后,通过回读对方执行程序的任务编号,达到双机同步控制,完成系统重构,实现对故障单元的不断电维修.通过对模拟机的实验表明,此控制系统能对运行中的错误实现实时的双机互检,且具有在线热插拔功能.

Fault-tolerance dual computer system based on maintainability without powering off

To improve the reliability of controller of a microcomputer, a fault-tolerant system which is consisted of two ones was proposed. In this system, output of a microcomputer was transmitted into another microcomputer through the serial ports, it was compared with the pre-stored data to carry out the diagnosis each other. The address encode technique was employed to implement the automatic switchover between two microcomputers and to insulate the fault unit diagnosed. After the fault unit was repaired and returned to the system(plug and play an electronic unit on-line), it waited through testing the task number of system program until the first command of next task was detached for synchronous. Experiment result shows that the controller can detect the fault in real time, and the failure device can be displaced without powering off.