COMPARISON OF TRADITIONAL AND NEURAL-NETWORK APPROACHES TO STOCHASTICNONLINEAR-SYSTEM IDENTIFICATION

A comparison between neural network and traditional approaches to nonl inear system identification is investigated with respect to aspects of model performance. Two neural network models, the state space and inp ut-output model structures, are considered. A global recurrent RMLP an d a leacher forcing RMLP are categorized as the stare space models, an d a global feedback FMLP and a teacher forcing FMLP are considered as the input-output models. In the traditional methods an AutoRegressive eXogeneous (ARX) Input model and a Nonlinear AutoRegressive eXogeneous (NARX) Input model are considered. Basic algorithms of models are des cribed, and simulation results are also presented through the system o utput response. Performance of models is compared based on the Mean-Sq uare-Errors (MSE). Noise-added sinusoidal, pulse and step signals are chosen as the test inputs for the validation of the obtained models. T wo different noise levels are augmented to the chosen input signals.