Identification of fractional order Wiener-Hammerstein systems based on adaptively fuzzy PSO and data filtering technique

This paper investigates the parameter estimation of fractional order Wiener-Hammerstein (FWH) nonlinear systems with colored noises. By employing data filtering, the original system with autoregressive moving average noise is filtered to the system with moving average noise; then, particle swarm optimization (PSO) is applied to identify the filtered system. To enhance the algorithm’s performance, the adaptively variable weight, dynamic learning factors and fuzzy control are introduced to construct the data filtering-based adaptively fuzzy PSO (DF-AFPSO) method. For a FWH system with known fractional order, DF-AFPSO is employed to identify the parameter vector, which consists of linear and nonlinear parameters. Furthermore, for a FWH system with unknown fractional order, DF-AFPSO can simultaneously estimate the parameter vector and fractional order by utilizing its parallel search ability. Finally, two simulation cases are designed to test the effectiveness of the proposed algorithm. The results illustrate that the DF-AFPSO method has higher accuracy in identifying FWH systems than the standard PSO and data filtering-based PSO methods.