Artificial Neural Network in the Autopilot System Application of Troubleshooting

Abstract

the existing fault diagnosis system in fault detection aspects of Boeing 737 A/P is effective, but in fault isolation aspects performance is poor, therefore using ANN technology need to improve its diagnosis system. A/P for the typical fault, the three layers feed forward artificial neural network structure, this paper introduces the conjugate gradient BP algorithm and gives the diagnosis results. Diagnosis results show that artificial neural network can accurately identify system three typical faults, improve the efficiency of fault diagnosis and fault isolation capability.