Study On The Autonomous Navigation System For An Omni-directional Four-wheeled Mobile Robot

Robots play a crucial role in the manufacture and assembly of products.Most modern manufacturing processes rely to a certain extent on robots and automated production lines.The most commonly used robot in the industrial environment is the robotic manipulator,which is made up of a movable arm fixed on the ground.This type of machine is controlled by a pre-programmed program and repeats some sequence of operations.They are usually faster,cheaper,and more accurate than human workers who perform the same task.A navigation system developed for an omni-directional wheeled mobile robot,called the Omnibot,is presented.This system is developed to enable the Omnibot to autonomously navigate,in a collision-free manner,along predefined paths in indoor structured office or factory-like environments.The navigation system is composed of four integrated subsystems: localization,path-following,velocity control,and obstacle detection.Firstly,execution of these motion commands is performed by the velocity control subsystem,which uses feedback control to regulate the angular velocities of the motors driving the Omnibot's wheels to produce the required motion of the robot.The kinematics and control algorithm of the robot are discussed,and its inverse kinematics model is proposed.Next,a localization system that uses a combination of odometry and a novel indoor GPS-like system provides the necessary estimates of the Omnibot's position and orientation(i.e.,pose).Using the pose updates from the localization subsystem,the path-following subsystem is able to compute motion commands to drive the Omnibot along the path.To ensure collision-free navigation,the Omnibot is equipped with an array of infrared distance sensors for detecting obstacles around its perimeter.The path-following subsystem is responsible for driving the Omnibot along a given path based on feedback about its location relative to its environment.Interaction between a human operator and the Omnibot is facilitated with a user-control interface running on a remote workstation.The interface allows the operator to visualize the Omnibot's location within a 3D model of its indoor workspace and provides a means to input commands.Finally,Testing of the developed system is performed,and the results confirm its effectiveness at enabling the Omnibot to perform collision-free autonomous navigation in an indoor structured environment.