Testing of two types of membership functions in fuzzy adaptive controller of pneumatic muscle actuator

摘要

Usage of pneumatic artificial muscles as actuators of manipulators is due to their highly nonlinear characteristics conditional by using the nonlinear control algorithms. Hybrid fuzzy adaptive control presented in the paper consists of a linear PD controller in the feedforward branch and a fuzzy adaptive controller with reference model in the adaptation branch and an output signal controlling the positioning system with two antagonistic artificial muscles is then created by multiplying the output signal from the PD controller and the adaptation gain of the fuzzy controller output. In the paper there are tested two types of membership functions for Sugeno fuzzy controller used in the adaptation subsystem. There is also shortly described metaheuristic method (genetic algorithm) used for parameters optimization of fuzzy controller.