Advanced search
1 file | 832.20 KB Add to list

Feedback control by online learning an inverse model

Author
Organization
Abstract
A model, predictor, or error estimator is often used by a feedback controller to control a plant. Creating such a model is difficult when the plant exhibits nonlinear behavior. In this paper, a novel online learning control framework is proposed that does not require explicit knowledge about the plant. This framework uses two learning modules, one for creating an inverse model, and the other for actually controlling the plant. Except for their inputs, they are identical. The inverse model learns by the exploration performed by the not yet fully trained controller, while the actual controller is based on the currently learned model. The proposed framework allows fast online learning of an accurate controller. The controller can be applied on a broad range of tasks with different dynamic characteristics. We validate this claim by applying our control framework on several control tasks: 1) the heating tank problem (slow nonlinear dynamics); 2) flight pitch control (slow linear dynamics); and 3) the balancing problem of a double inverted pendulum (fast linear and nonlinear dynamics). The results of these experiments show that fast learning and accurate control can be achieved. Furthermore, a comparison is made with some classical control approaches, and observations concerning convergence and stability are made.
Keywords
reservoir computing, pitch control, inverted pendulum, heating tank, neural network, feedback control, Adaptive control

Downloads

  • public.pdf
    • full text
    • |
    • open access
    • |
    • PDF
    • |
    • 832.20 KB

Citation

Please use this url to cite or link to this publication:

MLA
Waegeman, Tim, et al. “Feedback Control by Online Learning an Inverse Model.” IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS, vol. 23, no. 10, 2012, pp. 1637–48, doi:10.1109/TNNLS.2012.2208655.
APA
Waegeman, T., wyffels, F., & Schrauwen, B. (2012). Feedback control by online learning an inverse model. IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS, 23(10), 1637–1648. https://doi.org/10.1109/TNNLS.2012.2208655
Chicago author-date
Waegeman, Tim, Francis wyffels, and Benjamin Schrauwen. 2012. “Feedback Control by Online Learning an Inverse Model.” IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS 23 (10): 1637–48. https://doi.org/10.1109/TNNLS.2012.2208655.
Chicago author-date (all authors)
Waegeman, Tim, Francis wyffels, and Benjamin Schrauwen. 2012. “Feedback Control by Online Learning an Inverse Model.” IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS 23 (10): 1637–1648. doi:10.1109/TNNLS.2012.2208655.
Vancouver
1.
Waegeman T, wyffels F, Schrauwen B. Feedback control by online learning an inverse model. IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS. 2012;23(10):1637–48.
IEEE
[1]
T. Waegeman, F. wyffels, and B. Schrauwen, “Feedback control by online learning an inverse model,” IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS, vol. 23, no. 10, pp. 1637–1648, 2012.
@article{2986608,
  abstract     = {{A model, predictor, or error estimator is often used by a feedback controller to control a plant. Creating such a model is difficult when the plant exhibits nonlinear behavior. In this paper, a novel online learning control framework is proposed that does not require explicit knowledge about the plant. This framework uses two learning modules, one for creating an inverse model, and the other for actually controlling the plant. Except for their inputs, they are identical. The inverse model learns by the exploration performed by the not yet fully trained controller, while the actual controller is based on the currently learned model. The proposed framework allows fast online learning of an accurate controller. The controller can be applied on a broad range of tasks with different dynamic characteristics. We validate this claim by applying our control framework on several control tasks: 1) the heating tank problem (slow nonlinear dynamics); 2) flight pitch control (slow linear dynamics); and 3) the balancing problem of a double inverted pendulum (fast linear and nonlinear dynamics). The results of these experiments show that fast learning and accurate control can be achieved. Furthermore, a comparison is made with some classical control approaches, and observations concerning convergence and stability are made.}},
  author       = {{Waegeman, Tim and wyffels, Francis and Schrauwen, Benjamin}},
  issn         = {{2162-237X}},
  journal      = {{IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS}},
  keywords     = {{reservoir computing,pitch control,inverted pendulum,heating tank,neural network,feedback control,Adaptive control}},
  language     = {{eng}},
  number       = {{10}},
  pages        = {{1637--1648}},
  title        = {{Feedback control by online learning an inverse model}},
  url          = {{http://doi.org/10.1109/TNNLS.2012.2208655}},
  volume       = {{23}},
  year         = {{2012}},
}

Altmetric
View in Altmetric
Web of Science
Times cited: