A Nonlinear System Identification Method Based on Adaptive Neural Network

Nonlinear system identification (NSI) is of great significance to modern scientific engineering and control engineering. Despite their ability, the existing analysis methods for nonlinear systems have several limitations. The neural network (NN) can overcome some these limitations in NSI, but fail achieve desirable accuracy or training speed. This paper puts forward an NSI method based on adaptive NN, with aim further improve convergence speed NN-based NSI. Specifically, a generic model-based identifier was constructed, which integrates error feedback correction predictive model theory. Next, radial basis function (RBF) NN optimized by particle swarm optimization (PSO), used build model. effectiveness our were verified through experiments. research results provide reference applying PSO-optimized RBFNN other fields.