摘要
飞行器电子系统的多故障诊断问题可看作对各部件的故障概率求解的组合优化问题,在建立了系统故障诊断的概率因果模型基础上,运用遗传算法对飞行器电子设备各部件进行了故障定位。并针对遗传算法易陷入局优的缺点,提出了改进方法,将能量熵的选择加入到遗传算法的退火选择中,以充分地探索解空间,保持种群的多样性。试验结果表明,改进遗传算法能快速确定全局最优值,较好地解决了故障诊断领域中多故障关联的问题。
The multiple fault diagnosis for aircraft electronic system can be taken as a combination optimization problem to calculate the fault probability of every part. Based on building the probabilistic causal model for fault diagnosis, genetic algorithm(GA) is applied to aircraft electronic system fault location successfully. For GA is likely to fall into local optima, an improved GA is proposed. The energy-entropy selection is used in GA annealing selection, which can explore the solution space fully and keep the population diversity. The experimental results show that the improved GA is effective to determine the global optimal value quickly and solve the multiple-fault correlation problem in fault diagnosis.
出处
《系统工程与电子技术》
EI
CSCD
北大核心
2006年第5期788-790,共3页
Systems Engineering and Electronics
关键词
飞行器
电子系统
多故障诊断
遗传算法
aircraft
electronic system
multiple fault diagnosis
genetic algorithm