A calibration method for odometry of mobile robots based on the least-squares technique: theory and experimental validation

Simultaneous maximum-likelihood calibration of odometry and sensor parameters

“…This part of the method is numerically equivalent to the estimation done in [12]. The difference is that we are considering relative measurements instead of absolute, and very short intervals instead of full trajectories.…”

“…The method described in [12] estimates the matrix J as 4 independent numbers, here instead we use the three physical parameters (r L , r R , b) -this is actually harder, as there is one constraint more to consider. Anyway, their method could be easily extended to add one constraint to J, and our method could easily use independent elements.…”

“…In [11] the authors propose a maximum-likelihood model-free approach that does not seek a physical explanation of the odometry errors. In [12], the authors use a maximum-likelihood method for estimating the odometry parameters of a differential-drive robot, using the absolute observations of an external camera. The method is particularly simple because the problem is completely linear, and therefore can be solved via linear least-squares.…”

“…Compared to [12], we consider a three-parameters model for a differential-drive robot: we estimate the wheel radii r L , r R and the distance b between the wheels. For the sensor, we estimate its pose = ( x , y , θ ) with respect to the robot frame.…”

Simultaneous Parameter Calibration, Localization, and Mapping

“…In a similar way the odometry parameters can be estimated via another independent least-squares estimator given the knowledge of the reference trajectory [6]. For example, Antonelli et al [7,8] considered an external camera to track the position of a robot for calibrating the odometry parameters of a differential drive. Their method employs a least squares estimator which exploits the linear relation between the measurements and the unknown parameters.…”