Numerical and experimental investigation on the synthesis of extended Kalman filters for cable-driven parallel robots modeled through DAEs
Use this link to cite
http://hdl.handle.net/2183/35627
Except where otherwise noted, this item's license is described as Attribution 4.0 International ( CC BY 4.0)
Collections
- LIM - Artigos [52]
Metadata
Show full item recordTitle
Numerical and experimental investigation on the synthesis of extended Kalman filters for cable-driven parallel robots modeled through DAEsAuthor(s)
Date
2023-10-09Citation
Bettega, J., Boschetti, G., Frade, B.R. et al. Numerical and experimental investigation on the synthesis of extended Kalman filters for cable-driven parallel robots modeled through DAEs. Multibody Syst Dyn 60, 161–190 (2024). https://doi.org/10.1007/s11044-023-09941-5
Abstract
[Abstract]
Cable-driven parallel robots are parallel robots where light-weight cables replace rigid bodies to move an end-effector. Their peculiar design allows obtaining large workspaces, high dynamic handlings, ease of reconfigurability and, in general, low-cost architecture. Knowing the full state variables of a cable robot may be essential to implement advanced control and monitoring strategies and imposes the development of state observers. In this work a general approach to develop nonlinear state observers based on an extended Kalman filter (EKF) is proposed and validated both numerically and experimentally by referring to a cable-suspended parallel robot. The state observer is based on a system model obtained by converting a set of differential algebraic equations into ordinary differential equations through different formulations: the penalty formulation, the Udwadia–Kalaba formulation, and the Udwadia–Kalaba–Phohomsiri formulation, which have been chosen since they can handle the presence of redundant constraints as often happens in cable-driven parallel robots. In the numerical investigation, the EKF is validated simulating encoders heavily affected by quantization errors to demonstrate the filtering capabilities of EKF. In the experimental investigation, a very challenging validation is proposed: only two sensors measuring the rotations of two motors are used to estimate the actual position and velocity of the end-effector. This result cannot be achieved by sole forward kinematics and clearly proves the effectiveness of the proposed observer.
Keywords
State estimation
Extended Kalman filter
Cable-driven parallel robots
Penalty formulation
Udwadia-Kalaba
Udwadia–Kalaba–Phohomsiri
Estimación de estados
Filtro de Kalman extendido
Robots paralelos actuados por cable
Formulación en penalizadores
Extended Kalman filter
Cable-driven parallel robots
Penalty formulation
Udwadia-Kalaba
Udwadia–Kalaba–Phohomsiri
Estimación de estados
Filtro de Kalman extendido
Robots paralelos actuados por cable
Formulación en penalizadores
Editor version
Rights
Attribution 4.0 International ( CC BY 4.0)
ISSN
1573-272X