飞机交流发电机异常检测方法的研究  

Research on Method of Anomaly Detection of Aircraft AC Generator

作　　者：董恩生[1] 董永贵[1] 贾惠波[1] 

机构地区：[1]清华大学精密仪器与机械学系精密测试技术及仪器国家重点实验室,北京100084

出　　处：《仪器仪表学报》2004年第z1期153-154,共2页Chinese Journal of Scientific Instrument

摘　　要：飞机交流发电机是飞机电源系统中最重要的部件,其工作状态直接影响到飞行安全。在维护工作中,对发电机的输出信号进行异常检测,是发现发电机异常的常用手段,对及时排除故障隐患,保证飞行安全十分重要。借鉴计算机安全领域出现的人工免疫系统思想,采用欧氏距离在实数域进行匹配计算,实现了一种用于飞机交流发电机输出信号异常检测的负向选择算法。初步计算结果表明,该方法可有效检测出飞机交流发电机的输出异常。The aircraft AC generator is the most important component in the aero-power system.Its operation state directly affects the fligh security.Monitoring the output signal is a general way for finding the amomalous state of the generator.To trouble-shoot the latent fault and therefore guarantee the flight security,the anomaly detection is very important in the maintenance procedure.Referring to the artificial immune system idea appeared in computer security area,utilizing the Euclidean distance for matching calculation,a real-valued negative selection algorithm is implemented for the anomaly detection on the output signal of the aircraft AC generator. The initial results indicate that this algorithm can efficiently detect the anomaly of the output signal.

关 键 词：负向选择算法 异常检测 检测子 欧氏距离 

分 类 号：TH7-55[机械工程—仪器科学与技术]