Multi-Sensor Information Fusion for Mobile Robot Indoor-Outdoor Localization: A Zonotopic Set-Membership Estimation Approach

This paper investigates the localization of mobile robots in both indoor and outdoor scenarios. A zonotopic set-membership approach is proposed to fuse global navigation satellite system and odometry data outdoors, and 2D laser and odometry data indoors. Seamless switching between indoor and outdoor scene localization is achieved through a comparison of the current global navigation satellite system signal’s covariance with a predefined threshold in the proposed approach. Firstly, the robot’s position information is characterized using the odometry model, and the set containing the true state of the robot is updated to obtain the current updated zonotope. In addition, the global navigation satellite system or laser observation equations are described as a strip region and intersected with the prediction zonotope to obtain the feasible set of the states. Choosing the zonotope with the smallest volume from a family that encompasses the intersection of the two serves as the outer boundary for the intersection, enabling the determination of the precise position. The algorithm proposed in this paper can estimate the position state of the mobile robot to achieve accurate localization. To validate the proposed approach, relevant data are presented in the simulation results and discussion.