基于自主导航和深度学习的分拣服务机器人设计

Design of Sorting Service Robot Based on Autonomous Navigation and Deep Learning

为解决传统手动分拣方式效率低、错误率高且人力成本高昂的问题,对分拣服务机器人系统设计进行了研究。设计了一种基于自主导航和深度学习的分拣服务机器人,该系统采用改进A*算法优化了机器人在室内导航的全局路径规划,并通过YOLOv8算法与三维点云结合,实现了物体识别和三维定位,此外,手机APP-ROSClient可在局域网下通过Rosbridge与机器人系统进行数据交互,实现远程调度和监视。通过实验验证,该分拣服务机器人能够准确识别不同种类的货物,并在短时间内完成分拣任务。与传统的分拣方式相比,该机器人显著提高了分拣效率和准确性。该系统可实现智能化分拣过程,为服务机器人领域的研究提供了一种创新性解决方案,可应用于日常家庭、超市和酒店等场合,具有良好的经济和社会价值。

In response to the challenges posed by the inefficiency,high error rates,and elevated labor costs associated with traditional manual sorting methods,this study investigated the design of a sorting service robot system.A sorting service robot was conceptualized based on autonomous navigation and deep learning.The system optimized the global path planning for indoor robot navigation using an improved A*algorithm.Additionally,object recognition and three-dimensional positioning were achieved through the integration of the YOLOv8 algorithm with 3D point cloud data.Moreover,the ROSClient mobile app,connected to the robot system via Rosbridge over a local network,enabled remote scheduling and monitoring.Experimental validation confirmed the sorting service robot’s ability to accurately identify different types of goods and efficiently complete sorting tasks within a short time frame.In comparison to traditional sorting methods,the robot significantly improved sorting efficiency and accuracy.The system’s intelligent sorting process presents an innovative solution for the field of service robot research.Applicable in various settings such as households,supermarkets,and hotels,this system demonstrates significant economic and societal value.