Abstract—In many systems qualitative knowledge of the system behavior is sufficient for introducing a robust and adequately accurate response. This paper, presents a qualitative ant colony based algorithm to solve the model estimation problem in control of nonlinear dynamic systems. In this method, a qualitative model for dynamic systems is introduced as a graph and then the optimal path for shortest control time is obtained with the aid of ant colony system (ACS) optimization procedure. This method is well examined by a complicated two degree of freedom (2DOF) inverted pendulum system and the responses are presented. The presented approach has many advantages. First, it increases the robustness of the controlled system. Furthermore, it creates simple implementation through a single look up table (LUT). As a final point, it will be achieved control rule base generation for artificial intelligence methods such as fuzzy logic controller.
Index Terms— Ant colony system, Identification, Monte Carlo, Qualitative Control, Graph model.
A. Rezazade is with the Department of Electrical and Computer Engineering, Shahid Beheshti University, G. C, Evin, Tehran, Iran, 1983963113 (corresponding author to provide phone:+982177538160; fax:+982177538161;e-mail:a-rezazade@sbu.ac.ir, alireza.rezazade@gmail.com.)
[PDF]
Cite: Alireza Rezazade, "A Novel Qualitative Identification and Control of Two Degree of Freedom Inverted Pendulum based on Ant colony system,"
International Journal of Engineering and Technology vol. 1, no. 1, pp.13-20, 2009.
