现代电子技术
現代電子技術
현대전자기술
MODERN ELECTRONICS TECHNIQUE
2015年
1期
90-92
,共3页
故障诊断%FTA%BP神经网络%系统建模
故障診斷%FTA%BP神經網絡%繫統建模
고장진단%FTA%BP신경망락%계통건모
fault diagnosis%FTA%BP neural network%system modeling
针对某型电台的智能故障诊断，提出了故障树与BP神经网络相结合的诊断系统模型。BP神经网络的参数由故障树提取，具体诊断时，首先通过对发生的故障与故障实例库进行匹配，如果匹配成功，找出已有故障的维修策略，如果不匹配，则进入BP神经网络模块对新发生的故障进行诊断。结合这两种方法的电台故障诊断系统诊断速度快，准确率高，实用性强，在机载电子设备故障诊断中具有良好的应用前景。
針對某型電檯的智能故障診斷，提齣瞭故障樹與BP神經網絡相結閤的診斷繫統模型。BP神經網絡的參數由故障樹提取，具體診斷時，首先通過對髮生的故障與故障實例庫進行匹配，如果匹配成功，找齣已有故障的維脩策略，如果不匹配，則進入BP神經網絡模塊對新髮生的故障進行診斷。結閤這兩種方法的電檯故障診斷繫統診斷速度快，準確率高，實用性彊，在機載電子設備故障診斷中具有良好的應用前景。
침대모형전태적지능고장진단，제출료고장수여BP신경망락상결합적진단계통모형。BP신경망락적삼수유고장수제취，구체진단시，수선통과대발생적고장여고장실례고진행필배，여과필배성공，조출이유고장적유수책략，여과불필배，칙진입BP신경망락모괴대신발생적고장진행진단。결합저량충방법적전태고장진단계통진단속도쾌，준학솔고，실용성강，재궤재전자설비고장진단중구유량호적응용전경。
A fault diagnosis system model combining with fault tree and BP neural network is put forward for intelligent fault diagnosis of a certain radio station. The parameters of BP neural network are extracted by fault tree. Before diagnosing, the occurred fault is matched with fault case library. If matching is successful,try to find out a maintenance policy of the fault. Otherwise,enter BP neural network module to diagnose the faults. The diagnosis system combining with the two methods has high diagnosis speed,high accuracy and good practicability. It has a perfect application prospect in the fault diagnosis of air?borne electronic equipments.
