딥러닝을 이용한 무인 팔레트 트럭의 경로 제어 연구

초록

As e-commerce and pallet cargo business continue to grow and the skilled driver population decreases, the need for unmanned pallet trucks is increasing in the growing logistics unloading and loading works. Development of unmanned pallet trucks can be considered in two ways, including the development of unmanned fully automatic equipment and the method of unmanning existing manual pallet trucks using add-on technology. In this study, as a base technology for unmanned pallet trucks, research was conducted to develop key technology that can recognize standard pallets through cameras and deep learning recognition. Research was conducted to control the movement of the pallet truck in front of the pallet so that forks can be inserted more accurately and quickly into the pallet holes. As a result of this study, the accuracy of hole center recognition with standard pallet by YOLO deep learning from around 3 meters in front of the pallet was within ±10 mm and the performance was 10 Hz.