Ros2 ile sensör füzyonu kullanarak dinamik yol planlama ve haritalama

In this thesis, a path planning method is presented to ensure that mobile robots can safely and efficiently reach their targets in dynamic indoor environments where moving objects are present. This system, developed specifically for robots equipped solely with a LiDAR sensor, offers advantages in terms of cost and is suitable for real-time applications. During the path planning process, the robot uses LiDAR to detect obstacles in its surroundings and creates a dynamic environmental map, which is continuously updated to ensure safe navigation. In addition to LiDAR data, the robot also utilizes IMU (accelerometer, gyroscope, magnetometer) and odometry data to determine its position and orientation more accurately. These multiple data sources are combined through a process called sensor fusion in navigation. The developed system operates within the ROS2 framework, using the DWB (Dynamic Window Based) local planner integrated into the Nav2 architecture, and has shown successful results in environments with both static and dynamic obstacles. Simulations demonstrate that this sensor fusion-based approach improves the robot's motion stability and significantly reduces collisions.