Parameter estimation of Hammerstein systems based on the gravitational search algorithm

The Hammerstein nonlinear system is composed of a static nonlinear subsystem followed by a linear subsystem, and widely exists in chemical, thermal and other process industries. Therefore, it is significant to obtain parameter estimates of Hammerstein systems. This paper considers the identification problem of the Hammerstein finite impulse response moving average (FIR-MA) system using the gravitational search algorithm and the iterative identification technique. A particle swarm iterative identification algorithm is also proposed for comparison. The simulation results show that the proposed algorithms can effectively estimate the parameters of the Hammerstein FIR-MA system and the gravitational search algorithm behaves better than the particle swarm optimization approach in accuracy.