Single event upsets (SEU) are the transient errors that occur during the operation of the circuit. High radiation in the space environment and its invasion of the nanoelectronics can result in a bit-flip in the combinational circuits and may cause a stuck-at fault in the sequential circuit. Faults are unacceptable for any application, especially SEU in the control path, which is crucial and imperative since it can lead to functional or even mission failure. As a result, this paper proposes a bio-inspired technique based on a modified heuristic-guided genetic algorithm to mitigate error at the Finite State Machine (FSM), which is the controller’s behavior model. The proposed architecture performs an optimized functional level evolution of the FSM intrinsically on ProAsic3e FPGA boards without the help of System-on-Chip (SoC). Due to this, the delay caused by extrinsic and hybrid evolution has been reduced. The proposed heuristic-guided genetic algorithm recovers the fault in the control circuit with less convergence time when compared to the standard genetic algorithm. The resource utilization of the proposed evolvable hardware system has reduced costs compared to traditional functional evolution.

单事件翻转 (SEU) 是电路运行期间发生的瞬态错误。空间环境中的高辐射及其对纳米电子学的侵入会导致组合电路中的位翻转，并可能导致时序电路中的卡死故障。任何应用程序都不能接受故障，尤其是控制路径中的 SEU，这是至关重要且势在必行的，因为它可能导致功能甚至任务失败。因此，本文提出了一种基于改进的启发式引导遗传算法的仿生技术，以减轻控制器行为模型有限状态机 (FSM) 的误差。所提出的架构在没有片上系统 (SoC) 的帮助下，在 ProAsic3e FPGA 板上本质上执行 FSM 的优化功能级别演进。由于此，外部和混合进化造成的延迟已经减少。与标准遗传算法相比，所提出的启发式引导遗传算法以更短的收敛时间恢复了控制电路中的故障。与传统的功能演进相比，所提出的可演进硬件系统的资源利用率降低了成本。