On the Grey Dynamics of Type 2 Fuzzy Neural Hybrid Force Control

The problem of uncertainty for robot manipulator dynamic i n contact with an environment using hybrid force control and neuro-fuzzy is considered. Control of an industrial robot is mainly a problem of dynamics. It includes nonlinearity, uncertainties and external perturbations that should be considered in the design of control laws. For the first time, we formulate Type 2 neuro-fuzzy system based on extended back propagation (EBP) learning algorithm to adjust the parameters online, with no initial offline training while using the force error as the objective function. In the proposed method, a neural system is used as an approximate model of uncertain parts of robot dynamic while we assume by first principle knowledge there is known parts in robot dynamic. In addition, a self-tuning Type 2 fuzzy system is adopted to implement the on-line compensation for the static error caused by the PD controller based on the fuzzy rule set to improve the control performance. The proposed controller guarantees the closed loop stability for any arbitrary initial values of states and any unknown-but-bounded disturbances. Simulation results show the applicability and adaptability of the method to the hybrid force control and is more accurate when compared with alternative approaches.

On the Grey Dynamics of Type 2 Fuzzy Neural Hybrid Force Control Keywords:

Fuzzy system , Neural system , Robot , Uncertainty