实验科学与技术2026，Vol.24Issue(3)：12-19,8.DOI:10.12179/1672-4550.20240538
基于深度学习的智能象棋对弈机器人研究
Research on Intelligent Chess-Playing Robot Based on Deep Learning
摘要
Abstract
This research successfully developed an intelligent chess-playing robot using robotic arms based on Deep Q-Network,achieving innovations in machine vision,game algorithms,and mechanical structure.Based on YOLOv5 object detection technology and binocular stereo matching technology,the system achieves high-precision recognition of chess pieces(99%accuracy)and three-dimensional spatial positioning(error±1.2 mm).By integrating an improved minimax algorithm with Alpha-Beta pruning,the robot demonstrates high efficiency in game tree search and strategy evaluation.In addition,a multi-degree-of-freedom articulated robotic arm is employed to ensure high performance execution during gameplay.This research not only demonstrates the advancement of interdisciplinary integration in technology,but also exhibits efficient and reliable performance in practical applications,providing new insights for the automation and intelligent development of board games.Experimental results show that the chess-playing robot possesses high stability and precise recognition ability,ensuring accurate placement of each piece.It was also tested in games against human players,demonstrating excellent gameplay performance.关键词
深度学习/YOLOv5/机械臂/象棋/智能博弈Key words
deep learning/YOLOv5/robotic arm/Chinese chess/intelligent game playing分类
信息技术与安全科学引用本文复制引用
张攀..基于深度学习的智能象棋对弈机器人研究[J].实验科学与技术,2026,24(3):12-19,8.基金项目
北京科技大学教改基金项目(JC2020JY009) （JC2020JY009）
北京科技大学课程思政特色示范课建设项目(KC2023SZ46). （KC2023SZ46）
