Nonlinear system identification using Extended Possibilitic C-Means algorithm (EPCM) and Particle Swarm Optimization (PSO)

The Takagi-Sugeno fuzzy model is one of the best approaches for modeling and identifying of a nonlinear system. Several algorithms have been proposed in this framework; identify the premise parameters involved in the Takagi-Sugeno fuzzy model, as the fuzzy c-mean algorithm (FCM), the Gustafson Kessel algorithm (GK), PCM algorithm and EPCM algorithm. The implementation of these algorithms in the case of identification of nonlinear stochastic systems shows that this approach to several shortcomings, such as convergence to local optima and sensitivity to initialization (choice of number of clusters) and sensitivity at noise. In this paper, a combination of the EPCM algorithm and the PSO (particle swarm optimization) algorithm is used. However, the consequent parameters are therefore estimated by using the recursive weighted least squares (RWLS) method. The simulation results presented here illustrate the effectiveness of this algorithm.