NARX Deep Convolutional Fuzzy System for Modelling Nonlinear Dynamic Processes

This paper presents a new approach for modelling nonlinear dynamic processes (NDP). It is based on autoregressive with exogenous (NARX) inputs model structure and deep convolutional fuzzy system (DCFS). The DCFS hierarchical structure, which can overcome the deficiency of general systems when facing high dimensional data. For relieving curse dimensionality, as well improving approximation performance models, we propose combining NARX to provide good complex behavior fast-training algorithm ensured convergence. There are three structures proposed, appropriate training adapted. Evaluations were performed popular benchmark—Box Jenkin’s gas furnace data set four test systems. experiments show that proposed method be successfully used identify external dynamics static approximators.