Search

This article presents the implementation of the Unscented Kalman Filter (UKF) to improve the estimation of the pose of a differential mobile robot. The algorithm was validated and implemented using real robot data in MATLAB and Python, as well as its simulation in Gazebo-ROS. The results showed that the filter improved the estimation of the robot's state and performed well in all tests conducted. Additionally, the advantages of the UKF compared to other filters were discussed. The article also presented a control method for a differential robot that uses a PID on each wheel to follow a desired trajectory. It was concluded that the implementation of the UKF filter improved the estimation of the robot's pose. [Extracted from the article]

The article presents the implementation of autonomous navigation algorithms in a virtual simulation environment developed by the F1TENTH project. The selected algorithms include pure pursuit and a modification of the Braitenberg algorithm, which are combined to achieve desired behaviors. An improvement is proposed in the Braitenberg algorithm for high-speed obstacle avoidance. The algorithms were tested independently and in combination, achieving successful collision-free paths and desired trajectory tracking. The study was conducted by the CERLab at the University of Costa Rica and funded by the Vice-Rectorate of Research of the same university. The article titled "Global and Local Path Planning Study in a ROS-Based Research Platform for Autonomous Vehicles" presents a study on global and local route planning in a ROS-based research platform for autonomous vehicles. [Extracted from the article]