Nonlinear System Identification Based on GEP and GA Techniques

A method based on gene expression programming(GEP) for identifying the nonlinear system model is presented,which makes up the insufficiency that traditional identification methods need much a priori information,and has a tidier and more efficient system model expression mode than genetic programming(GP).It uses the improved genetic algorithm(GA) to carry out the model parameter evolution in a parallel mode,and the appropriate models can be obtained with limited given data.The definition of model fitness considers fully the accuracy and complicacy factors,and can get a trade-off identification solution.The simulation result indicates that the presented method can obtain nonlinear model in a quick and accurate way.