Neural Q-learning in motion planning for mobile robot

Author

Qin, Zheng ; Gu, Jason

Author_Institution

Dept. of Electr. & Comput. Eng., Dalhousie Univ., Halifax, NS, Canada

fYear

2009

fDate

5-7 Aug. 2009

Firstpage

1024

Lastpage

1028

Abstract

In order to solve the bad convergence property of neural network which is used to generalize reinforcement learning, the neural and case based Q-learning (NCQL) algorithm is proposed. The basic principle of NCQL is that the reinforcement learning is generalized by NN, and the convergence property and learning efficiency are promoted by cases. The detail elements of the learning algorithm are fulfilled according to the application of motion planning for mobile robot. The simulation results show the validility and practicability of the NCQL algorithm.

Keywords

convergence; learning (artificial intelligence); mobile robots; neurocontrollers; path planning; case based Q-learning algorithm; convergence property; learning efficiency; mobile robot; motion planning; neural network; reinforcement learning; Convergence; Interference; Learning; Logistics; Mobile robots; Motion planning; Multi-layer neural network; Neural networks; Robotics and automation; Sampling methods; Reinforcement learning; mobile robot; motion planning; neural network;

fLanguage

English

Publisher

ieee

Conference_Titel

Automation and Logistics, 2009. ICAL '09. IEEE International Conference on

Conference_Location

Shenyang

Print_ISBN

978-1-4244-4794-7

Electronic_ISBN

978-1-4244-4795-4

Type

conf

DOI

10.1109/ICAL.2009.5262570

Filename

5262570