Leader Following Control of Non-holonomic Mobile Robots Using EKF-based Localization

摘要

This paper presents a leader-follower controller combined with localization based on sensor data fusion. The leader robot has a cuboidal shape 3D printed tower on its top. ArUco markers with four different IDs are installed on the leader robot’s cuboidal tower. The follower robot has a one-degree-of-freedom rotating platform having Intel RealSense sensor on it. IMU sensors are installed on both the leader and follower robots. The data from wheel encoders and IMU sensors are fused by Extended Kalman Filter (EKF) to get the pose of robots. The proportional-integral-derivative (PID) controller rotates the Intel RealSense sensor on the follower robot to follow the ArUco markers on the leader robot. The follower robot calculates the pose of the leader robot after the detection of ArUco markers. Experiments are performed to validate the algorithm presented in the paper. Robot operating systems (ROS) is installed on both the robot’s single-board computers. The Opti-Track system is used to validate and plot the data fusion of robot poses.