Structure Identification and Adaptive Control of Neuro-Fuzzy Systems by a Non-parametric Regression Technique

摘要

The usefulness of control systems by neuro-fuzzy networks has been proved as an effective tool for non-linear systems. Determine the structure of a fuzzy system is not an easy task. In this paper, the initial structure of an adaptive neuro-fuzzy control system is determined by a non-parametric regression technique. This structure is used in conjunction with neural networks to deal with plant changes. The simulation results show the efficiency of this process.