摘要
通过分析直流电机的故障机理,得到在不同信号(如电流、转速、转矩等)中所表现的故障特征,提出了一种神经网络和D-S证据理论的多传感器数据融合技术的直流电机故障诊断方法。利用多源信息间的冗余性和互补性,有效提取故障特征信息,提高了诊断的可靠性和灵敏度。
A new method of fault diagnosis for DC machine which based on neural BP network and D-S Evidential Theory was proposing. After analyzed DC machine＇ s fault mechanism the fault features under vary signal such as electronic circuits rev and torque etc were obtained. Using redundancy and complementary of the multi-source signals, it selects features efficiency. The method was verified by simulation and the reliability and delicacy of diagnosis were improved.
出处
《电机与控制应用》
北大核心
2008年第2期49-51,64,共4页
Electric machines & control application
