Mobile Robot Motion Control Using Artificial Neural Net work Controll er

Abstract

Mobile robots are becoming more common in today's fast growing environment. Its extensive study and research have become a major part in the mobile robot's rapid development. An effective method of development is via modelling tools and computerized simulations. In this study, kinematic and dynamic models of a nonholonomic two-wheeled mobile robot were simulated with its behaviour. The robot defined in this study has two actuated wheels while any other contact with the surface travelled is assumed to be frictionless. This thesis, investigates how dynamics in recurrent neural networks canbe used to solve some specific mobile robot problems.A motion control approach based on a dynamic neural network has been designed. The advantage of this approach is that, no knowledge about the dynamic model is required, and no synaptic weight changing is needed in presence of time varying parameters. These control methods are implemented via the MATLAB/Simulink software into the kinematic and dynamic models of the robot. Controller was chosen according to its robot model that conforms to the standard robot designed in this study. The tracking control method of controller was also studied to ensure stability of the model. In the simulation, the robot is given several predetermined paths. The robot does not know these paths and it has to be able to adapt and react to different paths. The controller is considered successful when it can follow the predetermined path accordingly and effectively. The objective was to acquire a target, avoid obstacles and keep a geometric configuration at the same time.We have obtained successful results, on simulations.