Research on Fault Diagnosis of INS/GNSS/CNS Integrated Navigation System Based on GA-BP Neural Network

Aiming at the shortcomings of the current polar navigation, INS/GNSS/CNS integrated navigation system is pro-posed to improve the support ability of polar flight navigation. The fault diagnosis of INS/GNSS/CNS integrated navigation system is an important means to improve its reliability. The basic concept of BP neural network is introduced. The BP neural network optimized by genetic algorithm is raised for fault diagnosis of integrated navigation in view of the shortage that BP neural network is easy to sink into local minimum. The flow chart of GA-BP neural network and the structure model of fault diagnosis for integrated navigation are designed. The simulation test is carried out according to the algorithm and fault diagnosis model. The input parameters of BP neural network are optimized based on initial simulation. The simulation finally shows that the error between the output value and the expected value of GA-BP neural network is less than 1%, the simulation proves the reliability and accuracy of fault diagnosis for INS/GNSS/CNS integrated navigation system based GA-BP neural network.