Dynamics of flexible mobile manipulator structures.
Date
1997
Authors
Akpan, Unyime Okon.
Journal Title
Journal ISSN
Volume Title
Publisher
Dalhousie University
Abstract
Description
The topic of this thesis is the dynamics of mobile manipulators. To streamline the study and to facilitate systematic analysis mobile manipulators are divided into two broad categories: the wheeled and the non-wheeled systems. A non-wheeled mobile manipulator has a base which is much more massive than the manipulator structure and the dynamics of the manipulator does not affect the base dynamics. The wheeled mobile manipulator has a base mounted on wheels and the base dynamics is of the same order of magnitude as the manipulator and they are dynamically coupled with each other.
In non-wheeled mobile manipulators the base motion is modeled as a random process. The responses of the joints and the tip of the manipulator are studied as stationary and nonstationary random processes. Expressions for the covariance tensors of the Joint and the tip responses are developed. Single link and two-link manipulators are used to demonstrate the proposed analysis.
Two different models are used to study wheeled manipulators: the so-called quarter-car and the half-car models. The horizontal motion of the base is assumed, deterministic and known. The response of the wheeled manipulator system results from the motion on an irregular surface. The surface is modeled as a stochastic spatial field. Two cases of forward motion of the manipulator have been explored: uniform with constant speed and accelerated. The uniform forward motion produces purely stationary stochastic response while the accelerated motion produces nonstationary response in addition to a decaying deterministic component. Expressions for the covariance of the tip and the joints responses of the wheeled mobile manipulators are derived. Examples of two link manipulators mounted on quarter-car and half-car models are fully investigated.
In all mobile manipulator models studied in the thesis, the Singular Value Decomposition technique is used to derive expressions for the principal variance of the tip responses. The sensitivity of the principal variance of the tip responses to system parameters and configuration changes is investigated. It is shown that the principal variance of the tip motion is almost unidirectional and highly configuration dependent. To minimise the vibration of the manipulator tip it is suggested that: the "lower" links should be longer than the "upper" links; the damping efforts should be concentrated in the "lower" joints in addition to suspension damping; the "lower" joints should be stiffer than the "upper" joints. Therefore, it is suggested that for minimal tip vibration most of the design and control efforts should be focused on the "lower links and joints."
The Singular Value Decomposition technique is used to derive computation models for identification of the dynamic parameters of flexible mobile manipulators as well. Optimization criteria which can be used to set the manipulator configuration and excitation for efficient testing are proposed. Numerical simulations are discussed to validate the ideas.
Thesis (Ph.D.)--DalTech - Dalhousie University (Canada), 1997.
In non-wheeled mobile manipulators the base motion is modeled as a random process. The responses of the joints and the tip of the manipulator are studied as stationary and nonstationary random processes. Expressions for the covariance tensors of the Joint and the tip responses are developed. Single link and two-link manipulators are used to demonstrate the proposed analysis.
Two different models are used to study wheeled manipulators: the so-called quarter-car and the half-car models. The horizontal motion of the base is assumed, deterministic and known. The response of the wheeled manipulator system results from the motion on an irregular surface. The surface is modeled as a stochastic spatial field. Two cases of forward motion of the manipulator have been explored: uniform with constant speed and accelerated. The uniform forward motion produces purely stationary stochastic response while the accelerated motion produces nonstationary response in addition to a decaying deterministic component. Expressions for the covariance of the tip and the joints responses of the wheeled mobile manipulators are derived. Examples of two link manipulators mounted on quarter-car and half-car models are fully investigated.
In all mobile manipulator models studied in the thesis, the Singular Value Decomposition technique is used to derive expressions for the principal variance of the tip responses. The sensitivity of the principal variance of the tip responses to system parameters and configuration changes is investigated. It is shown that the principal variance of the tip motion is almost unidirectional and highly configuration dependent. To minimise the vibration of the manipulator tip it is suggested that: the "lower" links should be longer than the "upper" links; the damping efforts should be concentrated in the "lower" joints in addition to suspension damping; the "lower" joints should be stiffer than the "upper" joints. Therefore, it is suggested that for minimal tip vibration most of the design and control efforts should be focused on the "lower links and joints."
The Singular Value Decomposition technique is used to derive computation models for identification of the dynamic parameters of flexible mobile manipulators as well. Optimization criteria which can be used to set the manipulator configuration and excitation for efficient testing are proposed. Numerical simulations are discussed to validate the ideas.
Thesis (Ph.D.)--DalTech - Dalhousie University (Canada), 1997.
Keywords
Engineering, Mechanical.