Intelligent adaptive control of robotic dynamic systems with a newhybrid neuro-fuzzy-fractal approach

摘要

We describe a new method for adaptive model based control ofrobotic dynamic systems using a new neuro fuzzy fractal approach.Intelligent control of robotic systems is a difficult problem becausethe dynamic of these systems is highly nonlinear. We describe anintelligent system for controlling robot manipulators to illustrate ourneuro fuzzy fractal hybrid approach for adaptive control. We use a newfuzzy inference system for reasoning with multiple differentialequations for model selection based on the relevant selection parametersfor the problem. In this case, the fractal dimension of a time series ofmeasured values of the variables is used as a selection parameter. Weuse neural networks for identification and control of robotic dynamicsystems. The neural networks are trained with the Levenberg-Marquardt(LM) algorithm with real data to achieve the desired level ofperformance