시설물 감시 로봇의 경로계획법 연구

Abstract

Robots are being used for various purposes with the advancement of industry. Among them, mobile robots, which require three elements of perception, planning, and execution, are subject to active research because they can be widely applied to daily life and industrial environments. However, traditional mobile robots, using multiple ultrasonic sensors or infrared sensors, perceive the surrounding environment fragmentarily, thus facing limitations in creating a reliable global map due to significant errors in determining distances between obstacles and the robot in complex environments. To overcome this, recently, LiDAR has been developed to determine the distance to obstacles in a 2D or 3D range using multiple optical beams with a circular radius, and recognize complex shapes at high resolution. In addition, by utilizing IMU sensors, the physical status of the mobile robot can be assessed, enhancing the reliability of robot state estimation and creating a global map. In this study, we utilized these technologies to build an autonomous robot for monitoring facilities in Engineering Buildings 1 and 2 at Pukyong National University. This robot, equipped with a leveling device for stable driving on surfaces less even than indoors, was designed to perceive the robot and its surrounding environment through sensors such as LiDAR, IMU, and GPS. After experiments in virtual and real environments, we confirmed that the created facility monitoring robot reliably drives, transmits video, and flexibly plans and executes evasion routes for static or dynamic obstacles.