摘要
为解决航空发动机这一复杂系统的故障诊断问题,提高智能化诊断方法的准确率,使用了D-S证据理论对RBF神经网络、BP神经网络和支持向量机三个诊断子系统的诊断结果进行决策级融合,结果表明D-S证据理论的使用可以达到比单独运用三个子系统具有更好的诊断效能,经过融合降低了误诊率,改善了诊断性能。
The D-S evidence theory was used to improve the accuracy of intelligent diagnostic methods,which synchronized the result of the three diagnostic sub-system,the results showed that a better diagnostic performance can be achieved by using the D-S evidence theory than any others of the three separated sub-system,the ratio of misdiagnosis reduced and the diagnostic performance improved.
出处
《沈阳航空工业学院学报》
2010年第2期42-44,10,共4页
Journal of Shenyang Institute of Aeronautical Engineering