Item – Theses Canada

​
OCLC number
46528695
Author
Leng, David Yan,1964-
Title
Optimal trajectory planning for industrial robots.
Degree
M. Sc. -- University of Regina, 1994
Publisher
Ottawa : National Library of Canada = Bibliothèque nationale du Canada, 1994.
Description
2 microfiches.
Notes
Includes bibliographical references.
Abstract
As the number of industrial robots introduced into various manufacturing areas increases, requirements for their autonomy, flexibility and self-governing and ability to work in different environments become more severe. Hence, automatic trajectory planning of a manipulator in an environment with obstacles is a fundamental problem in robot design and control. The major effort of this thesis is to develop trajectory planning algorithms for robot manipulators considering collision avoidance. Two approaches, geometric and dynamic approaches, are developed for solving the problem. The geometric approach with two different methods is developed to achieve collision-free paths for robot movement. The first method is based on the shortest path planning. The path generated is the shortest and collision-free. The second method is based on cubic spline curves in 2-D space, which yields a collision-free, near shortest, single curve path. A numerical example is given using an IBM 7545 manipulator. Results are presented and discussed. The dynamic approach solves the trajectory planning problem using a simulation based method. In this approach, robot dynamics, singularity, workspace, and collision-free constraints are considered. A new collision avoidance algorithm is presented with only $O((V\sb R + M)F\sb P)$ complexity. Since the highly non-linear, non-differentiable, and non-convex constraints involved in this problem, a simulation method is used to solve this optimization problem. This simulation method is tested on a CRS-PLUS robot with a sphere obstacle in 3-D space. Numerical results and analysis are presented.
ISBN
031590996X
9780315909960