Modified Evolution Strategy Based Identification of Multi-input Single-Output Wiener-Hammerstein Model

Evolution strategies are a class of evolutionary algorithms with self-adaptation. A new method for identification of multi-input single-output (MISO) Wiener-Hammerstein model using evolution strategy is proposed. By using the least absolute residuals criterion, the identification scheme is cast as a complex nondifferentiable function optimization problem over parameter space. In order to find the optimal estimation of the system parameters, a modified evolution strategy is proposed to optimize the objective function. Simulation results show that the proposed method is a simple and effective non-recursive identification method, particularly for small samples.