Moonlit polarized skylight-aided INS/CNS: An enhanced attitude determination method

Autonomous navigation in complex environments (e.g., global navigation satellite system denied and unknown environments) is paramountly important for unmanned systems. Inspired by nocturnal insects’ navigation mechanism, an integrated navigation method combining the information of polarization navigation system (PNS), inertial navigation system (INS), and celestial navigation system (CNS) is presented for attitude determination at night. In order to improve the robustness of attitude estimation, a two-mode attitude determination system is designed, which includes the PNS/INS/CNS (PIC) mode and PNS/INS (PI) mode. With respect to the PIC mode, the bias in moon azimuth calculated by polarized skylight is modeled as a system state and subsequently estimated according to the precise starlight information. By detecting the CNS fault, the PIC mode is switched out. Focusing on PI mode, the estimated moon azimuth bias is applicable to improve the accuracy of the polarization information. A series of simulations and field tests are conducted to evaluate the performance of the presented method. The experimental data indicates that the presented method is able to achieve accurate and stable attitude estimation, which may offer valuable insight into the navigation of complex night environments.