Скачать 275.37 Kb.
robot trajectory planning IN
THE PRESENCE OF OBSTACLES
A thesis submitted in conformity with the requirements
for the Degree of Master of Applied Science
Graduate Department of Mechanical and Industrial Engineering
University of Toronto
© Copyright by Felix Wong, 2005
To my parents, Hilton and Susanna,
for their unceasingly love and care
without whose encouragement I would not have completed this thesis
I would like to express my sincere gratitude to Professor R. Ben Mrad and B. Benhabib for their help, guidance, and advice throughout the course of this thesis and my studies.
I am especially grateful to Tak Chan, for his work on the various electronic systems used in this thesis. I would also like to thank Imtehaze Heerah for assistance on the mechanical design of the mobile robot, Ardevan Bakhtari for his valuable discussions on my work, and Hor Wai Yeong, Marcia Hon, Vicky Wong, and Philip Abramson for their work on the experimental setup of this project.
I am also grateful for my colleagues and friends at the Computer Integrated Manufacturing Laboratory at the University of Toronto, in particular Farhad Agah, Hans Deruiter, Maurizio Ficocelli, Hamid Golmakani, Faraz Kunwar, Michael Naish, Goldie Nejat, William Owen, and Shichun Zhu. They have all been tremendously helpful with my work and have been a delight to be around.
I would like to thank my family for their love and support throughout all the years of my life. I am forever indebted to them. My gratitude also goes to Renee Lee for her love and encouragement. Lastly, I wish to thank the Lord, who has given me so much, and “who is good, whose love endures forever (Ps 118:29).”
Flexible Manufacturing Systems (FMS) constitute a popular alternative to traditional robotic manufacturing systems that require a high level of rigidity. An automated materials-handling system, such as Automated Guided Vehicles (AGVs), is a common component in an FMS. AGVs are currently designed to follow predefined paths embedded. Thus, their autonomy and efficiency can be enhanced if they could make on-line routing decisions. Two autonomy aspects are involved: obstacle avoidance and docking with a stationary or moving target. The objective of this thesis is to develop a novel on-line trajectory-planning method for fast robotic interception of moving objects in the presence of moving obstacles.
Numerous obstacle avoidance methods have been developed over the past two decades. They include Roadmap methods, Potential Field method, and Cell Decomposition methods. These methods consider the workspace in each instance as a static environment and find the shortest path to the target within their own constraints. Such a technique may work well under static conditions, but if the obstacles or target are not stationary, time-optimality quickly degrades. The proposed method, therefore, augments Cell Decomposition with the Proportional Navigation guidance law to achieve an algorithmically efficient and time-optimal solution.
The proposed trajectory planner has two components. The first tracks the target as if no obstacles are present. The second tracks a target point given by the Cell Decomposition method to avoid the obstacles in the workspace. Both of these components use Proportional Navigation to track its respective target. The time-to-impact with an obstacle is, then, estimated. Based on this value, a weight is assigned to the acceleration command of each component, depending on the necessity of obstacle avoidance. The final acceleration command is the summation of the two weighted acceleration commands and is executed by the robot.
The proposed trajectory planner is implemented in both computer simulations and experimentation. The performance was evaluated against the performance of conventional Cell Decomposition methods. The results show that the proposed hybrid trajectory planner is able to achieve tangibly faster interceptions.