Research of the Fogsins' Dynamic Error Model Based on Bp Neural Network
Yao Sun · Journal of Projectiles.Rockets.Missiles and Guidance · 2004
A new Fiber Optical Gyroscope Strapdown Inertial Navigation System's(FOGSINS)dynamic error model based on three layers Back-ropagation artificial neural network is developed.BP network model is trained with enough data-samples derived from orthogonal three axis-rate-input experiments,and tested by the real-time data.All training procedures of BP network model are performed by programming on Matlab.The simulation results prove 3-layers BP network has high precision,can be used as the dynamic error model in FOGSINS directly,and is possible to provide the mathematics foundation for the compensation of FOGSINS dynamic error.