Feedback linearization using Gaussian processes
- 1 December 2017
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 5249-5255
Abstract
Data-driven approaches from machine learning provide powerful tools to identify dynamical systems with limited prior knowledge of the model structure. This paper utilizes Gaussian processes, a Bayesian nonparametric approach, to learn a model for feedback linearization. By using a proper kernel structure, the training data for identification is collected while an existing controller runs the system. Using the identified dynamics, an improved controller, based on feedback linearization, is proposed. The analysis shows that the resulting system is globally uniformly ultimately bounded. We further derive a relationship between the training data of the system and the size of the ultimate bound to which the system converges with a certain probability. A simulation of a robotic system illustrates the proposed method.Keywords
This publication has 10 references indexed in Scilit:
- Safe learning of regions of attraction for uncertain, nonlinear systems with Gaussian processes2016
- Stability of Gaussian process state space models2016
- Safe and robust learning control with Gaussian processes2015
- Bayesian Nonparametric Adaptive Control Using Gaussian ProcessesIEEE Transactions on Neural Networks and Learning Systems, 2014
- Kernel methods in system identification, machine learning and function estimation: A surveyAutomatica, 2014
- Information-Theoretic Regret Bounds for Gaussian Process Optimization in the Bandit SettingIEEE Transactions on Information Theory, 2012
- Robust Nonlinear Control Associating Robust Feedback Linearization and$ H_infty$ControlIEEE Transactions on Automatic Control, 2006
- Dynamic systems identification with Gaussian processesMathematical and Computer Modelling of Dynamical Systems, 2005
- Feedback linearization using neural networksAutomatica, 1995
- Robust input-output feedback linearizationInternational Journal of Control, 1993