基于BP神经网络的虚拟故障诊断的实现

Implementation of Virtual Fault Diagnostics Based on BP Neural Network

针对电子设备故障诊断与维修训练的需求,以某型通信设备电源模块为例,探讨复杂电子设备元件级虚拟故障诊断的设计方案,重点研究了BP神经网络的实现、Delphi与Matlab的混合编程以及电路仿真在虚拟故障诊断中的应用,开发出某型通信设备的元件级虚拟故障诊断系统。实践证明,该系统实现了故障仿真和诊断以及虚拟拆装演示,具有较好的交互性,能够很好地解决虚拟维修过程中遇到的实际问题。

Aiming at the requirement of electronic equipments’fault diagnosing and maintenance training,taking the power sources of certain type of communication equipment as an example,the design scheme of component-level virtual fault diagnostics system of complex electronic equipments is discussed,and the realization of BP neural network,mixed programming of matlab and Delphi and the application of the circuit simulation in virtual fault diagnostics are researched,and the component-level virtual fault diagnostics system of certain type of communication equipment is developed.The practical results show that this system realizes faults simulation,diagnostics and virtual dismantling with preferable interactions,which solves the problems effectively in the process of virtual maintenance.