Novel hybrid of strong tracking Kalman filter and wavelet neural network for GPS/INS during GPS outages

• A novel INS errors prediction model based on INS errors trend was proposed.
• A novel combination of STKF/WNN algorithm was proposed to maintain GPS/INS performance.
• Different INS errors prediction models were compared and analyzed.
• Different AI-based algorithms were compared and analyzed.
• Simulation and test results indicated the effectiveness of the proposed method.

Aiming to improve positioning precision of the GPS/INS integrated navigation system during GPS outages, a novel model combined with strong tracking Kalman filter (STKF) and wavelet neural network (WNN) algorithms for INS errors compensation is proposed and tested. STKF is used to estimate INS errors as a replacement of Kalman filter (KF), and WNN is applied to establish a highly accurate model based on STKF when GPS works well and to predict INS errors during GPS outages. Performance of the proposed model has been experimentally verified using GPS and INS data collected in a land vehicle navigation test. The comparison results indicate that the proposed model combined with STKF/WNN algorithms can effectively provide high accurate corrections to the standalone INS during GPS outages.