Simultaneous Calibration And Mapping For Mobile Robot With Non-Holonomic Constraint

In this paper, a simultaneous calibration and mapping algorithm for mobile robot is proposed. The concept of expectation maximization (EM) are applied. Two main processes, mapping and calibration, are executed iteratively until converged. Graph SLAM are applied in mapping. For calibration, an innovative algorithm is proposed to decouple the whole problem into several sub-steps. And non-holonomic constraint of the mobile robot kinematics is utilized to speed up the computation. Experiment on an Autonomous Guided Vehicle (AGV) system are conducted to validate our algorithm.