DocumentCode :
952166
Title :
A cartpole experiment benchmark for trainable controllers
Author :
Geva, Shlomo ; Sitte, Joaquin
Author_Institution :
Sch. of Comput. Sci., Queensland Univ. of Technol., Brisbane, Qld., Australia
Volume :
13
Issue :
5
fYear :
1993
Firstpage :
40
Lastpage :
51
Abstract :
The inverted pendulum problem, i.e., the cartpole, which is often used for demonstrating the success of neural network learning methods, is addressed. It is shown that a random search in weight space can quickly uncover coefficients (weights) for controllers that work over a wide range of initial conditions. This result indicates that success in finding a satisfactory neural controller is not sufficient proof for the effectiveness of unsupervised training methods. By analyzing the dynamics of the linear controller, the cartpole problem is reformulated to make it a more stringent test for neural training methods. A review of the literature on unsupervised training methods for cartpole controllers shows that the published results are difficult to compare and that for most of the methods there is not clear evidence of better performance than the random search method.<>
Keywords :
adaptive control; nonlinear control systems; cartpole experiment benchmark; controller weights; inverted pendulum problem; neural network learning methods; trainable controllers; Artificial neural networks; Computer simulation; Control systems; Control theory; Delay; Humans; Neural networks; Search methods; Testing; Weight control;
fLanguage :
English
Journal_Title :
Control Systems, IEEE
Publisher :
ieee
ISSN :
1066-033X
Type :
jour
DOI :
10.1109/37.236324
Filename :
236324
Link To Document :
بازگشت