摘要
应用多传感器信息融合理论与不确定性决策理论,研究了基于神经网络和D-S证据理论相结合的信息融合的故障诊断方法。以传感器检测数据为输入,以神经网络输出构造基本概率赋值函数,对不同传感器的检测证据按D-S证据理论进行融合,得到待识别目标的诊断可信度。通过对柴油机振动监测数据、燃油压力波动信息以及两者融合信息的故障诊断结果的比较,表明基于神经网络和D-S证据理论相结合的多传感器信息融合方法用于复杂机械的故障诊断是可行和有效的。
Adopting the multi-sensor information fusion theory and indeterminate decision-making theory,the failure diagnosis method based on the information fusion which combines neural network and D-S evidence theory was studied.With sensor detection data as input and neural network output constituting the basic probability assignment function,the detection evidences of different sensors were fused according to the D-S theory to obtain the reliability of the object to be identified.By comparing diesel engine vibration monitoring data,fuel pressure fluctuation information and the failure diagnosis result of their fused information,it is shown that application of the multi-sensor information fusion method based on neural network and D-S evidence theory is feasible and effective to complex mechanical failure diagnosis.
出处
《石油机械》
北大核心
2010年第6期49-52,72,共5页
China Petroleum Machinery
基金
陕西省自然科学基金项目"石油钻井过程安全预警与多源信息融合智能监控技术研究"(2006E12)
陕西省教育厅专项科研计划项目"基于信息融合的钻井过程事故智能监控与预警技术"(07JK365)
关键词
多传感器
信息融合
神经网络
D—S证据理论故障诊断
multi-sensor,information fusion,neural network,D-S evidence theory,failure diagnosis
