双处理器间容错控制技术

Fault-tolerant control technology for double CPUs

针对双处理器的协同处理问题,文中提出一种基于FPGA的容错控制技术,并将其应用于双处理器系统共管模式。将处理器及其相关资源的故障分为4种情况进行讨论,综合运用故障模式响应、状态监控和控制、动态故障判断、仲裁控制和资源重配置等方法实现故障的检测、屏蔽和切换。该方案具有体积小、功耗低、可靠性高的特点。最后,以人为干预的形式采取相应的物理手段,分别向模块注入外部故障,通过计算系统切换时间来验证容错控制器的切换成功率和切换效率,并分析其对系统的影响。

In allusion to the problem of CPU(central processing unit) cooperative processing,a fault-tolerant control technology based on FPGA is proposed,and applied to the common management mode of double-CPU system. The faults of CPU and its related resources are divided into four cases for discussion. The fault mode response,state monitoring and control,dynamic fault judgment,arbitration control and resource reconfiguration are used to realize fault detection,shielding and switching. The scheme has the characteristics of small volume,low power consumption and high reliability. The corresponding physical means are applied by means of human intervention to inject the external faults into the module respectively. The switching success rate and switching efficiency of the fault-tolerant controller are verified by calculating the system switching time,and its impact on the system is analyzed.