Pso based on the key-term separation principle for fractional hammerstein system identification

The block oriented structure such as Hammerstein, Wiener, etc... became very popular for the non-linear system modelling due to its simplicity and parsimony. In this paper, fractional hammerstein system is considered; it consists of a static non-linear block followed by a fractional linear dynamical block. The particle swarm optimisation is used to estimate the system parameters as well as the fractional order based on the key term separation principle. It allows the transfer of the original Hammerstein model into a new regression identification model, this way, the redundant parameters estimation is avoided. Different simulations test the method efficiency at various signal to noise ratios.