Design of the Self-Constructing Fuzzy Neural Network controller for a sliding door system

In this paper, the Self-Constructing Fuzzy Neural Network (SCFNN) controller suitable for real-time control of the speed control of the slide door is presented to track reference model. The structure and parameter learning can be done automatically and online. The structure learning is accordance with the partition of input space (error and change of error), and the parameter learning is based on the supervised gradient decent method. In this paper, the weights of SCFNN are generated from functional-link-based neural network (FLNN). The SCFNN adopted the FLNN, generating complex nonlinear combinations of input space to the weights of the SCFNN with FLNN. Finally, a slide door speed control system is implemented in this paper to verify the effectiveness of the proposed SCFNN with FLNN.