Feedback-based iterative learning control for constrained systems with application to robotic manipulators
Document TypePhD thesis
Access StatusOpen Access
© 2019 Dr. Gijo Sebastian
Iterative learning control (ILC) is an advanced control algorithm that achieves tracking of the desired reference trajectory through repetitions without using the precise knowledge of the dynamic systems. In a feed-forward ILC algorithm, the tracking errors from the past trials are utilized for learning the control input. Even though the method is simple to implement and analyse, the transient error in the iteration-domain is usually not regulated. In a feedback-based ILC, a feedback (or tracking error of current trial) is incorporated with a feed-forward ILC to improve the transient behaviour as well as providing more design freedom and robustness to non-repeatable disturbances. When the system of interest is nonlinear with input and output constraints, the convergence analysis of ILC becomes more complex due to the existence of feedback in the control structure. The thesis investigates how these constraints can be handled in such a feedback-based ILC scheme with a limited model information. The thesis proposes novel learning control architectures and analysis methods to separately handle input constraints and output constraints for a class of nonlinear dynamic systems with rigorous convergence analysis. The contributions of this thesis are presented in two parts. The first part proposes a novel structure that can handle input constraints in a feedback-based ILC system in a systematic manner. The second part proposes a barrier function based feedback design in ILC that can ensure the satisfaction of output constraints during the learning process. For simplicity and consistency of the thesis, the focus here is on the continuous-time plant model. The main contributions of the thesis are summarised as follows: 1. When dealing with input constraints for a feedback-based ILC, a new composite energy function is proposed to ensure the convergence and boundedness of trajectories. This new CEF can be used to unify the two design methods: Contraction Mapping and Composite Energy Function. 2. Barrier-Lyapunov function is employed to address hard output constraints in a feedback-based ILC scheme. 3. The proposed algorithms have been tested on a robotic manipulator platform. The simulation and experimental results have shown that the proposed feedback based ILC can handle input and output constraints with limited knowledge of the system. Hence the proposed methods can be applied to a large class of engineering systems when the precise models are hard to obtain.
Keywordsiterative learning control; robotic manipulators; constrained systems
- Click on "Export Reference in RIS Format" and choose "open with... Endnote".
- Click on "Export Reference in RIS Format". Login to Refworks, go to References => Import References