飞控系统传感器故障诊断的神经网络方法研究

Research of Neural Network on Sensor Fault Diagnosis for Flight Control System

BP神经网络对于飞行控制系统传感器故障诊断是一种有效的故障模式识别方法;在标准BP神经网络的基础上,提出了一种新的BP改进算法——自适应FMBP算法(SAFMBP),用以消除标准BP网络收敛速度慢及易陷入局部极小等缺点,并且建立了飞行控制系统仿真模型和传感器常见故障模型,采用基于神经网络模式分类的故障诊断方法,应用改进的BP神经网络(SAFMBP)进行飞控系统传感器的故障诊断,最后给出了仿真诊断实例。

Back-propagation （BP） neural network algorithm is one of effective pattern recognition methods for the sensor fault diagnosis of flight control system （FCS）. In this paper, we propose a new improved BP algorithm-Self-Adaptive FM BP algorithm （SAFMBP） based on the cOnventional BP network, which can solve the problems of slow convergence speed and easily getting into local minimum in the conventional BP algorithm. With the establishment of models of FCS and the corresponding sensor faults, the improved BP neural network is applied in the FCS sensor fault diagnosis via the fault diagnosis approach of neural network pattern classification. Simulation results are given.