Study of cooperative position estimations of mobile robots

This work presents a study to determine the improvement of the simultaneous position estimations of several Mobile Robots (MRs) adding the relative distances between them, regarding independent position estimations for each MR. To evaluate the performance of the proposal, it is supposed that the localization scenario is composed of an Ultrasonic Local Positioning System (ULPS) and two MRs with odometry information that describe different trajectories. The independent position estimations of the MRs are obtained using an Extended Kalman Filter (EKF) for each MR that fuses the ULPS data and the odometry information. To evaluate the improvement in the estimation of the MR positions when the information of the relative distance between them is available, another EKF that estimates at the same time the positions of all MRs is implemented. The proposal has been validated by simulations, demonstrating that the simultaneous position estimations of both MRs using one EKF with the relative distance information presents an improvement in comparison with the independent estimations. Furthermore, the value of this improvement has been quantified considering the quality of the distance measurements between the MRs, supposing different noise levels based on a Gaussian distribution. Finally, a complexity evaluation of the independent and cooperative estimations is shown in terms of number of operations (products and additions) for each iteration of the EKF implementations.