基于神经网络的某装备电路单元故障诊断研究

ANN-Based Fault Diagnosis Method of Circuit Units for a Certain Weapon Equipment

某装备是一种较为先进的防空武器系统.该系统结构和电路复杂,元器件种类与品种多,技术保障中其性能检测与故障诊断的难度大.为此,针对系统电路单元功能模块的组成情况,进行了单元电路模块化诊断方法研究.以该装备某单元电路板为例,介绍了利用神经网络和模式识别相结合,实现功能模块的故障判断并对故障元器件定位的方法.结果表明:这种方法能够大大降低测前仿真的工作量,提高故障诊断的精度和速度,增强该武器装备的维修保障能力.

Because a certain aerial defense weapon system consists of large amounts of components and complex circuits, it is difficult to check the performance and diagnose the fault in technology support. In accordance with the actual circuit units composed by function modules, and taking the position regulator board as an example, we proposed a module diagnosis method. Neural network and pattern recognition were combined to realize the fault diagnose of the functional modules and the fault component location. Result shows that this method can greatly lower the artificial work load, and improve the precision and pace of fault diagnosis.