Dynamic identification of robots with power model

Abstract

This paper presents a new approach to identify the minimum dynamic parameters of robots using least squares techniques (LS) and a power model. Theoretical analysis is carried out from a filtering point of view and clearly shows the superiority of the power model over the energy one and over the dynamic identification model which has been used to carry out a classical ordinary LS estimation and a new weighted LS estimation. These results are checked from comparing experimental identification of the dynamic parameters of a planar scara prototype robot.

Nonlinear Friction and Dynamical Identification for a Robot Manipulator with Improved Cuckoo Search Algorithm

“…Then, Grotjahn and Daemi [20] proposed an interpolated trajectory consisting of two parts, part I overcame the given boundary conditions and part II overcame the homogeneous boundary conditions. Furthermore, Gautier [21] used fifth-order polynomials to obtain smooth joint trajectories to be executed by the manipulator, and the polynomial coefficients are fixed by imposing continuity constraints between the trajectory segments. In this paper, we adopt a finite Fourier series which was proposed by Swevers et al [22] as excitation trajectories.…”

Identification of robot dynamic parameters using Jacobi differentiator

“…However, this method can amplify the noise effect in the estimations ofq,q. To avoid the noise distribution, (q,q,q) must be filtered by a low-pass filter F q (s), with derivative operator s. The filter F q (s) should have a flat amplitude without phase shift in the range [0 ω cq ], with the rule of thumb ω cq > (10 × ω dyn ), and ω dyn is the bandwidth of the joint position closed loop [24]. Meanwhile, the torque Γ m is perturbed by high frequency torque ripple from joint drive chain in the closed loop control.…”

“…The most widely applied approach is based on the robot explicit dynamic model, requiring joint acceleration data which are usually estimated from noisy measurement [22]. The other approaches are based on the robot energy model [23] or the robot power model [24], which require only joint velocity data, but instead they need a derivation operation on an implicit part of velocity. Besides, some authors utilize a parallel scheme to identify robot dynamic parameters by minimizing the output error from a closed loop simulation [25].…”

Motion Planning Taking Into Account the Dynamic Model of Vehicles: Application to the Compactor

“…Using this property of the Inverse Dynamic Model, a Weighted Least Squares method of identification is proposed by (Gautier, 1997) to obtain the values of the dynamic parameters X s of robot manipulators and applied to a mobile machine, the compactor, in (Guillo et al, 1999).…”