Target Tracking Robot Design Based On ROS Environment

With the rapid development of science and technology,the applications of intelligent robot in production and life become more and more widespread.Among these applications,intelligent robots based on visual tracking technology play an important role in fields such as supermarket shopping,airport baggage handling and factory material distribution.Taking supermarket shopping and airport baggage handling as the application scene,this thesis designs a robot with target tracking function based on ROS environment.Firstly,the structural modules of the target tracking robot are designed,including the division of the functions,the construction of the hardware circuit and software programming.Secondly,the tracking method of target tracking robot is researched.The basic principle of Camshift algorithm and multi-feature fusion Camshift algorithm is analyzed.Based on the above,this thesis proposes an improved multi-feature fusion Camshift algorithm.By building the ROS environment,the tracking algorithm is applied to the robot to carry on the target tracking experiment,which verifies the tracking performance of the improved multi-feature fusion Camshift algorithm.Finally,the control method in robot trajectory tracking is studied.The kinematic model of the target tracking robot is established based on its kinetic characteristics.Considering that the motion control system of mobile robot includes two inputs and three outputs as a typical underactuated system,the sliding mode control method is applied to track the robot’s motion trajectory.The trajectory tracking of a robot includes position tracking and attitude tracking in which the position tracking is internal loop tracking and the attitude tracking is external loop tracking.The sliding surface and control law are designed according to the robot kinematics model and motion control requirements,and the robot’s position and attitude are effectively controlled.The simulation results verify the robustness and reliability of the sliding mode control algorithm.