A hybrid prediction method and its application in the distributed low-cost INS/GPS integrated navigation system

In order to improve the accuracy of INS/GPS integrated navigation system during GPS signals blockage, an effective and low-cost method is to design the corresponding linear or non-linear predictor to predict the position and velocity errors between INS and GPS during GPS blockage and then to correct the results of INS. Based on the distributed data fusion system, a novel hybrid prediction method that combines the radial basis function network (RBFN) and Kalman filter (KF) together was proposed. The predicted value is divided into two parts. One part is the innovation component of KF and the other is the state prediction component of KF. The former is predicted with the designed 6 RBFNs; the latter is predicted with two distributed KFs. Through practical experiments and data processes, it is shown that the proposed hybrid predictor possibly improve the accuracy of INS during GPS blockage.