Nonlinear system identification using new Extended Possibilistic C-Means Algorithm and Particle Swarm Optimization
Troudi Ahmed, Lassad Houcine, Bouzbida Mohamed, Abdelkader Châari · 2013
The development of a mathematical model making it possible to represent ˝ as well as possible ˝ the dynamic behaviors of a complex real process represents a very important problem in the real world. Fuzzy logic and more, particularly the Takagi-Sugeno (TS) fuzzy model draws the attention of several researchers during these last decades. This is due to their capability to approximate the nonlinear system in several locally linear subsystems. Many clustering algorithms exist in literature allowing the identification of the parameters intervening in the TS fuzzy model. In this paper a new clustering algorithm noted NEPCM-PSO is proposed. The proposed algorithm represents a combination between New Extended Possibilistic C-Means algorithm (NEPCM) and Particle Swarm Optimization (PSO) algorithm. The effectiveness of this algorithm is tested on a nonlinear system and on an electro-hydraulic system. In this paper a comparative study between PCM algorithm, NEPCM algorithm and NEPCM-PSO algorithm are also presented.