Motion control of flexible arm by using neuro-adaptive control: LQN control system and numerical simulation

Abstract

In this article, a new control method for rotational motion control of a flexible arm by using an LQN control system is proposed. This control system is based on combining discrete-time state feedback control by linear quadratic optimal control theory with neuro-adaptive feedforward control by a neural network model. While the discrete-time state feedback control loop provides the elastic vibration control, the neuro-adaptive feedforward control loop controls the endpoint position. Numerical simulation is carried out for a single flexible arm by employing the proposed LQN control system. From the simulation results, it is confirmed that the elastic vibration in positioning control is effectively suppressed, and control error of the endpoint position decreases gradually with the passing of time, and in steady-state, the control error becomes almost zero. Consequently, desired endpoint motion is achieved, showing the effectiveness of the proposed control method by use of the LQN control system.