VIO/GPS Fusion Positioning Algorithm of Substation Inspection Robot Based on EKF

With the development of the electric power industry, intelligent inspection systems for substations have been widely applied to ensure the safe operation of substations and maintain power supply quality. To address the issue of low localization accuracy in existing substation inspection robot navigation technologies, this paper proposes a fusion algorithm based on visual-inertial odometry and GPS using Extended Kalman Filtering (EKF). The algorithm propagates the filter state through visual-inertial estimation and updates it using GPS measurements, effectively eliminating cumulative drift in position and heading, while requiring minimal computational resources. Theoretical and experimental results demonstrate that this algorithm can achieve globally consistent and accurate six-degree-of-freedom pose estimation in outdoor substation environments.