基于二级信息融合的装备故障诊断方法

Equipment fault diagnosis method based on two step data fusion

针对武器装备故障诊断的多故障问题和单一诊断方法推理能力弱、匹配能力差的缺点,采用RBF神经网络数据级融合和DS证据理论特征级融合相结合的方法应用在二级故障诊断模型中,给出了模型实现步骤,并结合某型导弹制导电子箱故障诊断进行了实验验证。该法使诊断不确定性大大减小,克服了单一方法的缺陷与不足,并使武器装备故障诊断的准确度得到提高。

The two step fault diagnosis model of RBF neural network data fusion and DS evidence theory characteristic fusion is given to solve the multiple faults problem in weapons and equipment fault diagnosis and overcome the defects of a single diagnostic method＇s weakly reasoning and matching capacity.The step to realize the model is given.The model was verified by experiments combined with a certain type of missile guidance electronic box failure,and it reduced the diagnosis uncertainty greatly.This method overcomes the shortcomings and deficiencies of a single method and improves the fault diagnosis accuracy of the weapons and equipment.