A high-accuracy constrained SINS/CNS tight integrated navigation for high-orbit automated transfer vehicles. (October 2018)

• Record Type: Journal Article
• Authors: Wang, Dingjie; Lv, Hanfeng; An, Xueying; Wu, Jie
• Is Part Of: Acta astronautica. Volume 151(2018)
• Page Start: 614
• Page End: 625
• Publication Date: 2018-10
• Subjects: Strapdown inertial navigation system (SINS) -- Tight integration -- Model constraint -- Automated transfer vehicle (ATV); Astronautics -- Periodicals; Outer space -- Exploration -- Periodicals; Astronautics; Periodicals; 629.405
• DOI: 10.1016/j.actaastro.2018.07.015
• Languages: English
• ISSNs: 0094-5765

Abstract

Abstract: High-accuracy and reliable autonomous navigation is increasingly crucial for automated transfer vehicles (ATV). This paper proposes a novel strapdown inertial navigation system/celestial navigation system (SINS/CNS) tight integration scheme aided by dynamic model constraints for high-orbit ATV to realize accurate and autonomous navigation. In this scheme, the complete weightlessness constraint in orbit is used to address the divergence of position and velocity caused by inaccurate accelerometer bias estimation problem encountered in the traditional SINS/CNS integration method, and the image point position-based tight integration model is derived to handle the adverse influence of time-varying attitude measurement noise due to changes of star geometry observed by a large-view-filed star sensor. Moreover, an information filter is devised to fuse the multi-rate measurements. The proposed algorithm is evaluated by a representative high-orbit ATV trajectory simulation, which indicates significant improvements in navigation accuracy compared with its traditional counterparts. The proposed algorithm can realize navigation accuracy enhancements without introducing additional sensors, strengthening its potentials in engineering application. Highlights: Accelerometer-bias and star geometry undermine traditional SINS/CNS integration. Accurate bias estimation can be achieved using the complete weightlessness in orbit. Changing star geometry can be tackled using image-point-based tight model. The proposed algorithm improves SINS/CNS navigation accuracy remarkably.