A federated filtering integrated navigation system with dynamic step gradient descent algorithm was designed to reduce the performance of UAV attitude solution tracking under strong maneuvering and high speed motion. SINS/GNSS/CNS/PA federal filter was used to estimate the motion acceleration, and then the step size of gradient descent method was dynamically adjusted to suppress the influence of motion acceleration on the aircraft attitude solution. At the same time, the fault detection mechanism based on residual estimation can prevent information pollution caused by satellite signal jump or starlight occlusion in high speed maneuvering environment. The simulation results show that the proposed method improves the attitude calculation accuracy and enhances the filtering accuracy of the UAV integrated navigation system in complex motion state compared with the commonly used integrated navigation filter.
Research on SINS/GNSS/CNS/Pressure Altimeter Federated Filtering Adaptive Fault Tolerance Method
Lect. Notes Electrical Eng.
China Satellite Navigation Conference ; 2022 ; Beijing, China May 22, 2022 - May 25, 2022
China Satellite Navigation Conference (CSNC 2022) Proceedings ; Chapter : 38 ; 444-454
2022-05-10
11 pages
Article/Chapter (Book)
Electronic Resource
English
Research on Barometric Altimeter Assisting GNSS Navigation for Urban Environments
Springer Verlag | 2018
|Research on Barometric Altimeter Assisting GNSS Navigation for Urban Environments
British Library Conference Proceedings | 2018
|