Talebi, Heidar Ali (1997) Neural network-based control of flexible-link manipulators. PhD thesis, Concordia University.
The problem of modeling and control of flexible-link manipulators has received much attention in the past several years. There are a number of potential advantages arising from the use of light-weight flexible-link manipulators, such as faster operation, lower energy consumption, and higher load-carrying capacity. However, structural flexibility causes many difficulties in modeling the manipulator dynamics and guaranteeing stable and efficient motion of the manipulator end-effector. Control difficulties are mainly due to the non-colocated nature of the sensor and actuator position, which results in unstable zero dynamics. Further complications arise because of the highly nonlinear nature of the system and the difficulty involved in accurately modeling various friction and backlash terms. Control strategies that ignore these problems generally fail to provide satisfactory closed-loop performance. This dissertation presents experimental evaluation on the performance of neural network-based controllers for tip position tracking of flexible-link manipulators. The controllers are designed by utilizing the output redefinition approach to overcome the problem caused by the non-minimum phase characteristic of the flexible-link system. Four different neural network schemes are proposed. The first two schemes are developed by using a modified version of the "feedback-error-learning" approach to learn the inverse dynamics of the flexible manipulator. The neural networks are trained and employed as online controllers. Both schemes require only a linear model of the system for defining the new outputs and for designing conventional PD-type controllers. This assumption is relaxed in the third and fourth schemes. In the third scheme, the controller is designed based on tracking the hub position while controlling the elastic deflection at the tip. In the fourth scheme which employs two neural networks, the first network (referred to as the output neural network) is responsible for specifying an appropriate output for ensuring minimum phase behavior of the system. The second neural network is responsible for implementing an inverse dynamics controller. Both networks are trained online. Finally, the four proposed neural network controllers are implemented oil a single flexible-link experimental test-bed. Experimental and simulation results are presented to illustrate the advantages and improved performance of the proposed tip position tracking controllers over the conventional PD-type controllers in the presence of unmodeled dynamics such as hub friction and stiction and payload variations.
|Divisions:||Concordia University > Faculty of Engineering and Computer Science > Electrical and Computer Engineering|
|Item Type:||Thesis (PhD)|
|Authors:||Talebi, Heidar Ali|
|Pagination:||xix, 149 leaves : ill. ; 29 cm.|
|Degree Name:||Theses (Ph.D.)|
|Program:||Electrical and Computer Engineering|
|Thesis Supervisor(s):||Khorasani, Khashayar|
|Deposited By:||Concordia University Libraries|
|Deposited On:||27 Aug 2009 13:11|
|Last Modified:||08 Dec 2010 10:14|
Repository Staff Only: item control page