Computer Player for Gomoku and Pick-and-Place Task Using Robotic Arm

• Main Authors: SU, JING-YI, 蘇靖壹
• Other Authors: CHANG, CHIH-HUNG
• Format: Others
• Language: en_US
• Published: 2018
• Online Access: http://ndltd.ncl.edu.tw/handle/3yg9fz

碩士 === 國立高雄大學 === 應用數學系碩博士班 === 107 === Deep learning and intelligent manufacturing are widely applied nowadays. In this paper, we implement two applications of these two fields. The application of deep learning is to train a computer player for Gomoku. Gomoku is a chess game with a simple rule that whoever gets their stones 5-in-a-row wins. We first train the Gomoku player using supervised learning. Secondly, we also enhance our Gomoku player with reinforcement learning. The application of intelligent manufacturing is to accomplish a pick-and-place task using a robotic arm. Instead of hard coding for specific object and target positions, we want the robotic arm to detect the positions and accomplish the task automatically. So far, some results have been achieved.