This paper proposes an image-based visual servoing of an omnidirectional mobile robot without velocity sensor using a multi-layer artificial neural network dynamics. The multi-layer artificial neural network dynamics is trained with the actual input and output data of the omnidirectional mobile robot so that it can represent the dynamic characteristics of the reference dynamics well. On the one hand, the velocity sensors attached to the actuators of the omnidirectional mobile robot contain uncertainties. Therefore, one of the applications of the trained dynamics is to implement image-based visual servoing of an omnidirectional mobile robot without velocity sensor. the simulation results are provided to verify the validity of the proposed method.
