Comparison of Least-Squares and Instrumental Variables for Parameters Estimation on Differential Drive Mobile Robots. Issue 7 (2021)

• Record Type: Journal Article
• Authors: Ardiani, Fabio; Benoussaad, Mourad; Janot, Alexandre
• Is Part Of: IFAC-PapersOnLine. Volume 54:Issue 7(2021)
• Page Start: 310
• Page End: 315
• Publication Date: 2021
• Subjects: Identification -- Parameter Estimation -- Mobile Robots -- Dynamic Modelling; Automatic control -- Periodicals; 629.805
• DOI: 10.1016/j.ifacol.2021.08.377
• Languages: English
• ISSNs: 2405-8963

Abstract

Abstract: This paper addresses the parameter estimation issue on mobile robots. A comparison between the state-of-art Least-Squares Technique and the potentially useful Instrumental Variable Method is carried out. With that objective, the whole process of kinematic and dynamic modeling, exciting trajectory design, simulation, parameter identification and cross-validation is done. The results are exhibit in simulation of a Differential Drive Mobile Robot to show that against possible noises and perturbations on mobile robotics, due to effects as slipping and not perfect rolling, Instrumental Variable is more suitable. As a use case, a mobile robot with an unbalanced mass distribution is modeled, identified and validated in simulation.