基于神经网络的余度敏感器结构故障诊断方法

Fault Diagnosis for Redundancy Sensitive Elements Based on Artificial Neural Network

针对飞行载体中余度敏感器结构的故障诊断,讨论了基于神经网络的直接比较测量值的诊断方法。应用自适应线性元件能实现线性,感知器具有分类的特点,设计出一种神经网络,可用于余度敏感器结构的故障诊断。仿真结果表明:这种方法不仅可以检测出故障、故障的类型和发生的时间,而且灵敏度高,为余度敏感器结构的故障诊断提出了一种有实用价值的方法。

Discusses a new diagnosis approach to directly and compares the measurement based on ANN,for the fault diagnosis of redundancy sensitive elements in the carrier.Utilizing the high linear property of the adaptive linear element and classification of perceptrons,designed a neural network,which can be used in the fault diagnosis of redundancy sensitive elements.Simulation results show that the method not only can detect fault,fault type and occurring time,but also is sensitive to faults.So this new approach is useful and practicable.