Combined Navigation System Based on Improved Neural Network Enhanced cubature kalman filter

Combined navigation systems based on microelectromechanical systems (MEMS) and Global Positioning System (GPS) with inertial devices lose satellite signals under harsh conditions, resulting in rapid deviations from the pure inertial navigation positioning errors. To solve this problem, a combined navigation algorithm based on improved Particle Swarm Optimisation (PSO) algorithm and Mean Value Filtering (MVF) optimised Radial Basis Function Neural Network (RBFNN) to augment the CKF algorithm is proposed, and singular value decomposition (SVD) is introduced to improve the stability of the filtering updata. The use of a trained neural network is used to correct the error of an inertial guidance system in case of GPS signal loss. Finally, the performance of the method is verified by on-board simulation experiments. The experimental results show that the method can suppress effectively the error dispersion of the system in the case of GPS signal loss.