Abstract ▼ AI-Helper

Due to the recent development of autonomous robot technology, various robots are being used in real life. Among them, mobile robots perform tasks such as moving goods, cleaning and guiding in service environments such as warehouses, hotels, and restaurants. A mobile robot uses an algorithm that gene...

Due to the recent development of autonomous robot technology, various robots are being used in real life. Among them, mobile robots perform tasks such as moving goods, cleaning and guiding in service environments such as warehouses, hotels, and restaurants. A mobile robot uses an algorithm that generates a path from a given environment to a destination in order to perform a mission. This is called Path Planning, and Path Planning is divided into two stages: global Path Planning(GPP) and local Path Planning(LPP). GPP mainly generates the shortest distance to the destination avoiding obstacles in a simplified environment, and LPP creates a path that takes into account the dynamic constraints of the robot and complex dynamic and static obstacle avoidance within the result of GPP.
In this thesis, a Path Planning method was developed that optimizes the Bézier curve following the any-angle Path Planning as GPP through Spiral-Optimization (SPO) and Golden Standard Search (GSS). To verify this, the any-angle Path Planning method and the optimization-based Bézier curve generation method were simulated in the Matlab environment, and the skid bot was controlled on the simulation that simulated the real situation using the Robot Operating System (ROS) and GAZEBO. An actual experiment was also performed using a motion capture system and a steer bot, and compared with the GSS-only Bézier curve generation method. As a result of simulation and actual experiment, any-angle Path Planning was able to quickly generate a path from the map of the given experimental environment, and based on the path planning result, a smooth path could be obtained through optimization of the Bézier curve. Finally, it was confirmed that it can move to the destination by better avoiding obstacles in the experimental environment when the Optimization-Based Bézier curve generation technique using SPO-GSS is used rather than when the robot moves along the curve using GSS alone.