Estimation of Object Kinematics From Point Data

Abstract

One of the fundamental problems arising in kinematics is that of determining object position, velocity and acceleration from given point position, velocity and acceleration data. This type of problem is frequently encountered in robotics, biomechanics, real-time control of space structures, automatic guided vehicles, etc. Complications arise when redundant data are used and when the data have errors. Chutakanonta and Gupta proposed two simple and elegant methods for the estimation of object position from the given point position data. The present work is an extension of these methods for estimating the object velocity and acceleration states from the given point position, velocity and acceleration data. The method proposed herein uses Singular Value Decomposition (SVD) to effectively estimate the object velocity and acceleration states. Such matrix decompositions can be performed by using readily available matrix-oriented software like MATLAB and can be successfully used to simplify the solution of the over-determined system of equations encountered in these types of problems. Several hypothetical examples and examples that simulate practical situations are presented to determine the effectiveness, robustness and applicability of the proposed method. The method is found to be very effective in estimating the object velocity and acceleration states in the presence of imprecise and redundant data as well as for nearly co-planar point data.

A model predictive approach for online mobile manipulation of non-holonomic objects using learned dynamics

“…The estimation of dynamic parameters of a manipulated object by an autonomous mobile robot has received some attention in the past (Stüber et al, 2020). Most of the existing approaches either require an extensive training dataset (Fitzpatrick et al, 2003) or use kinematics-based methods for a specific task (Vithani and Gupta, 2002).…”

“…Previous research concerning dynamics models mostly describe either mapping between actions and consequences for a specific task (Fitzpatrick et al, 2003; Ogata et al, 2005) or rely on pure kinematics (Vithani and Gupta, 2002). A more advanced approach mimics human sensorimotor learning behavior, in which a coarse dynamics model of the new object is learned based upon prior beliefs and experiences.…”

Application of vector finite screw analysis to determine the roll centre from wheel points

“…Screw axis analysis is widely used in the fields of robotics and biomechanics. Several methods have been developed to determine the screw parameters given the measurements of the motion of landmarks on a rigid body [4][5][6][7][8][9]. In contrast to the non-linear analysis of finite displacements using rotation matrix, velocity analysis is a linear problem with simple explicit equations.…”