基于循环神经网络的飞机电源健康状态自动监测方法  

Automatic Monitoring Method for Aircraft Power Supply Health Status Based on Recurrent Neural Network

作　　者：郑煜昕 ZHENG Yuxin(Shenyang Aircraft Industry(Group)Co.,Ltd.,Shenyang 110000,China)

机构地区：[1]沈阳飞机工业(集团)有限公司,辽宁沈阳110000

出　　处：《通信电源技术》2024年第17期110-112,共3页Telecom Power Technology

摘　　要：由于影响飞机电源健康状态的因素较多,且具体的影响程度不同,对飞机电源健康状态的监测结果往往存在偏差。为此,提出基于循环神经网络的飞机电源健康状态自动监测方法。考虑故障是引起飞机电源健康状态波动的主要因素,构建了以飞机电源状态为核心,包含主电源失效故障、备用电源失效故障以及配电系统故障的故障树。在飞机电源健康状态监测阶段,引入循环神经网络,计算故障树底事件对飞机电源健康状态的影响权重,并综合所有存在故障的影响程度,确定最终的健康状态。在测试结果中,设计方法在飞机电源健康状态监测中展现出显著优越性,可以精准监测飞机电源健康状态。Due to the numerous factors that affect the health status of aircraft power supplies and their varying degrees of impact,there are often deviations in the monitoring results of aircraft power supply health status.Therefore,an automatic monitoring method for aircraft power supply health status based on recurrent neural networks is proposed.Considering that faults are the main factor causing fluctuations in the health status of aircraft power supplies,a fault tree was constructed with the aircraft power supply status as the core,including main power supply failure faults,backup power supply failure faults,and distribution system faults.In the stage of monitoring the health status of aircraft power supply,a recurrent neural network is introduced to calculate the weight of the impact of fault tree bottom events on the health status of aircraft power supply,and comprehensively evaluate the degree of impact of all existing faults to determine the final health status.In the test results,the design method demonstrated significant superiority in monitoring the health status of aircraft power supplies,enabling accurate monitoring of the health status of aircraft power supplies.

关 键 词：循环神经网络 飞机电源健康状态 自动监测 故障树 影响权重 

分 类 号：TM63[电气工程—电力系统及自动化]