Multi-Sensor Fusion Technology in Inertial Navigation System Using Factor Graph

This paper presents a novel algorithm that is able to satisfy the demands of multi-sensor fusion. It uses a probabilistic model based on factor graph, which provides a better understanding of navigation. solution on non-linear optimization. This architecture which achieves plug-and-play can receive multi-rate, asynchronous measurements. The incoming measurements from various sensors are added to the factor graph as nodes and form an entire graph structure. We design a total state estimator including velocity and poses of the vehicle. To reduce the angular error because of gyroscope drift, a novel quaternion-based on attitude estimator with magnetic, angular rate and gravity (MARC) sensor arrays are used to ensure the accuracy of attitude heading reference system (AHRS). The attitude estimator provides a better reference value for the inertial navigation system in time. The simulation results show that the approach not only provides a solution of high accuracy and reliability, but also meets the need of various condition using the multi-sensor fusion.