基于LS-SVM的雷达电子装备故障诊断研究  

作　　者：朱家元[1] 宋宁哲[1] 姜朝毅[1] 

机构地区：[1]空军雷达学院陆基预警监视装备系,武汉

出　　处：《21世纪（理论实践探索）》2010年第3期125-127,共3页The tuenty-first century

摘　　要：采用智能技术进行电子装备故障诊断是故障诊断领域的重点研究课题。该文研究了新型最小二乘支持向量机（LS—SVM），设计了多元分类最小二乘支持向量机算法，在此基础上，建立了电子装备智能故障诊断模型，并对某型雷达电子装备进行了故障诊断。结果表明，该智能诊断模型具有出色的学习分类能力和泛化能力，故障诊断准确，具有良好的工程应用价值。That using intelligent methods for electronic equipment fault diagnosis is one of core research fields in fault diagnosis. A new least squares support vector machines is studied in the paper, and a improved multi-class least squares support vector machines algorithm is present. And then, a electronic equipment systems intelligent fault diagnosis mode is designed based on the new least squares support vector machines algorithm. Ferthermore, the new mode is used for fault diagnosis of radar electronic equipment systems. The results show that the new intelligent fault diagnosis mode is feasible which has more valid learning generalization ability comparing to traditional methods, and also it has great engineering application value.

关 键 词：神经网络 支持向量机 雷达 电子装备 故障诊断 

分 类 号：TN97[电子电信—信号与信息处理]