Neuro-controller using simultaneous perturbation for a flexible arm system

This paper describes a neuro-controller system to control a flexible arm. The simultaneous perturbation method is adopted as a learning rule of the neuro-controller. Using a direct inverse control scheme by a neural network; we control the flexible arm. Then the neural network must learn an inverse system of the flexible arm. If we use the gradient method as a learning rule of the neural network Jacobian of the plant is required. On the other hand, our control scheme described here does not require information about the plant such as Jacobian, because the simultaneous perturbation method estimates the gradient using only values of the error defined by output of the plant and its desired one. Simulation results and actual control results of a real flexible beam system are described to confirm the feasibility of the proposed method