SLAM/Calibration of Multiple LiDARs Using Nonlinear Observers: Algorithm and Simulation

This paper considers a nonlinear observer method for robust SLAM (simultaneous localization and mapping) with online calibration parameter estimation. Our design contains the design of a continuous-time pose/landmark observer in a cascaded interconnection with a robust discrete-time calibration parameter estimator. Both the nonlinear observers are potential to possess certain stability as well as appealing robustness in a sense of input-to-state stability (ISS) like characterizations. The proposed results are also illustrated by extensive numerical simulations.