Skip to Main Content
In neural modeling of non-linear dynamic systems, the neural inputs may include any system inputs of interest and system outputs with various time delays. To obtain the optimal subset of inputs regarding a performance measure is a combinational problem, and the selection process can be time-consuming. In this paper, the input selection is transformed to a model structure selection problem and a new input selection method is proposed. This method is then applied to the neural modeling of a practical system, and the modeling result shows its merit.