论文部分内容阅读
针对现有故障诊断方法难以诊断涵盖多种不同类型故障的问题,提出一种基于分层DSmT的多故障诊断方法.利用主元凝聚层聚类方法实现证据聚类,将辨识框架分成若干个子框架;利用证据主元将BP神经网络所生成的各种故障模式的基本概率赋值函数在不同辨识框架下重新分配;利用DSmT对子框架下的证据进行融合并得出诊断结果.仿真实验结果表明,所提出的方法能将不同类型故障从辨识框架中分离出来,提高多故障诊断结果的可靠性,减少计算量,提高诊断效率.
Aiming at the problem that existing fault diagnosis methods are difficult to diagnose and cover many different types of faults, a multi-fault diagnosis method based on hierarchical DSmT is proposed. The principal component cohesion layer clustering method is used to implement evidence clustering. The recognition framework is divided into several sub-frames ; The basic probability assignment function of various failure modes generated by BP neural network is re-distributed under different identification frameworks by using evidence principal components; the evidence under the sub-framework is fused by DSmT and the diagnosis results are obtained.The simulation results show that, The proposed method can separate different types of faults from the identification framework, improve the reliability of multi-fault diagnosis, reduce the computational load and improve the diagnostic efficiency.