Nonlinear system identification with genetic algorithms

Based on the genetic algorithm and Volterra series model, a new identification method of nonlinear systems is proposed. The method is adapted to the difference magnitude of parametric of model. An adaptation algorithm is developed in order to get a wider search range and achieve more accurate solutions synchronously. With unique mutation and selection, an effective genetic algorithm is achieved for solving the identification of nonlinear systems. The simulation and experimental results indicate that the new method achieved high accuracy, robustness and adaptability