Abstract
An approximate dynamic programming (ADP) based supplementary learning control method is developed to online improve the performance of existing controllers. The proposed supplementary learning structure can make full use of the prior knowledge of the pre-designed controller and endow the controller with learning ability. Moreover, by introducing the action dependent value function for policy evaluation, the supplementary learning control can work in a model-free manner. The policy iteration algorithm is employed to train the actor-critic structure of the ADP supplementary controller. Simulation studies are carried out on the cart-pole system to validate the optimization and the adaptation capability of the proposed methodology.
| Original language | English (US) |
|---|---|
| Title of host publication | 26th Chinese Control and Decision Conference, CCDC 2014 |
| Publisher | IEEE Computer Society |
| Pages | 1316-1321 |
| Number of pages | 6 |
| ISBN (Print) | 9781479937066 |
| DOIs | |
| State | Published - 2014 |
| Event | 26th Chinese Control and Decision Conference, CCDC 2014 - Changsha, China Duration: May 31 2014 → Jun 2 2014 |
Other
| Other | 26th Chinese Control and Decision Conference, CCDC 2014 |
|---|---|
| Country | China |
| City | Changsha |
| Period | 5/31/14 → 6/2/14 |
Fingerprint
Keywords
- Approximate Dynamic Programming
- Model-Free
- Online
- Supplementary Control
ASJC Scopus subject areas
- Information Systems and Management
- Control and Systems Engineering
Cite this
Online and model-free supplementary learning control based on approximate dynamic programming. / Guo, Wentao; Liu, Feng; Si, Jennie; Mei, Shengwei.
26th Chinese Control and Decision Conference, CCDC 2014. IEEE Computer Society, 2014. p. 1316-1321 6852370.Research output: Chapter in Book/Report/Conference proceeding › Conference contribution
}
TY - GEN
T1 - Online and model-free supplementary learning control based on approximate dynamic programming
AU - Guo, Wentao
AU - Liu, Feng
AU - Si, Jennie
AU - Mei, Shengwei
PY - 2014
Y1 - 2014
N2 - An approximate dynamic programming (ADP) based supplementary learning control method is developed to online improve the performance of existing controllers. The proposed supplementary learning structure can make full use of the prior knowledge of the pre-designed controller and endow the controller with learning ability. Moreover, by introducing the action dependent value function for policy evaluation, the supplementary learning control can work in a model-free manner. The policy iteration algorithm is employed to train the actor-critic structure of the ADP supplementary controller. Simulation studies are carried out on the cart-pole system to validate the optimization and the adaptation capability of the proposed methodology.
AB - An approximate dynamic programming (ADP) based supplementary learning control method is developed to online improve the performance of existing controllers. The proposed supplementary learning structure can make full use of the prior knowledge of the pre-designed controller and endow the controller with learning ability. Moreover, by introducing the action dependent value function for policy evaluation, the supplementary learning control can work in a model-free manner. The policy iteration algorithm is employed to train the actor-critic structure of the ADP supplementary controller. Simulation studies are carried out on the cart-pole system to validate the optimization and the adaptation capability of the proposed methodology.
KW - Approximate Dynamic Programming
KW - Model-Free
KW - Online
KW - Supplementary Control
UR - http://www.scopus.com/inward/record.url?scp=84905242181&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84905242181&partnerID=8YFLogxK
U2 - 10.1109/CCDC.2014.6852370
DO - 10.1109/CCDC.2014.6852370
M3 - Conference contribution
AN - SCOPUS:84905242181
SN - 9781479937066
SP - 1316
EP - 1321
BT - 26th Chinese Control and Decision Conference, CCDC 2014
PB - IEEE Computer Society
ER -