Farnaz Sabahi - Control Engineering, Shahrood University of Technology, Shahrood, Iran
Mohamad Mehdi Fateh - Department of Control Engineering, Faculty of Electrical & Robotic Engineering, Shahrood University of Technology, Shahrood, Iran
Ali Akbar Gharevici - Department of Control Engineering, Faculty of Electrical & Robotic Engineering, Shahrood University of Technology, Shahrood, Iran

The problem of uncertainty for robot manipulator dynamic in contact with an environment using impedance control and neural network is considered. Control of an industrial robot is mainly a problem of dynamics. It includes nonlinearities, uncertainties and external perturbations that should be considered in the design of control laws.In the proposed method a perceptron network is used as an approximator for modeling uncertain parts of robot dynamic but we assumed by first principle knowledge there is known parts in robot dynamic.Extended backpropagation learning algorithm is used to adjust the parameters of network.Neural network parameter matrices are adapted online ,with no initial offline training, using the force error as the objective function. The neuro-controller guarantees the closed loop stability for any arbitrary initial values of states, neural network parameters and any unknown-but-bounded disturbances.Simulation results show the applicatibility and adaptability of the method to the impedance force control.

Neural Network, Impedance Control , Robot, Uncertainty