Research On Multi-Sensor Data Fusion Applying To Vehicle Integrated Navigation System

In this paper, according to working environment and application requirements of vehicle integrated navigation system, it takes GPS/DR vehicle integrated navigation system as the application object, the federated Kalman filter was designed and the simulation experiment was carried out, as well, it introduces a fuzzy Kalman filter algorithm, by simulation the feasibility and effect was proved.Firstly, it studied Kalman filter, which is one of the multi-sensor data fusion algorithm; and then the Federated Kalman filter for vehicle GPS/DR integrated navigation system was designed, the federated Kalman filter equation was developed, simulation and analysis was carried out, by theory and experience results, we can find that the fault-tolerant and real-time of distributed filter is better, which met the system requirements; finally, fuzzy Kalman filter algorithm was introduced in this system, the simulation shows that through on-line adjustment of the measurement noise matrix, the algorithm can suppress filter divergent-factors and improve anti-interference ability, robustness effectively, it can adapt the change of noise environments.