摘要
阐述了自动测试系统中故障诊断的重要性,详述了诊断的流程。以某型防空导弹自动驾驶仪为例,首先利用LabVIEW中的MATLAB脚本节点,分别以BP和RBF神经网络为算法编写故障诊断函数;然后用LabVIEW开发故障诊断模块界面,最后将LabVIEW和神经网络两者的优势结合实现自动驾驶仪的故障诊断。实验仿真结果表明:模块设计合理,人机互动界面友好,神经网络训练效果较好,模块能较好地完成故障诊断功能。修改诊断模块的部分程序后可将其移植到其他自动测试系统,模块的设计流程以及诊断函数具有一定的通用性和参考价值。
The importance and flow of fault diagnosis in automatic test system( ATS ) are presented. In the case of autopilot, the algorithms based on BP and RBF neural network function respectively are programmed to diagnose fault in the form of MATLAB scripts in LabVIEW. Then LabVIEW serves as a programming tool to develop a fault diagnosis interface for an ATS. The last step is to integrate scripts into LabVIEW codes f or fault diagnosis of autopilot. The results of practical experiment show that the diagnosis interface of module is friendly and the neural network is trained well as well as the fault diagnosis can be achieved successfully. It notes that the proposed programming codes, fault diagnosis flow and module embrace universal characteristics can be applied to other ATSs.
出处
《航天控制》
CSCD
北大核心
2011年第5期58-62,76,共6页
Aerospace Control