文献类型：期刊文献

中文题名：基于云的餐厅服务机器人系统设计

英文题名：Design of Restaurant Service Robot System Based on Cloud

作者：王博玮[1];陆中成[1]

机构：[1]华东理工大学信息科学与工程学院

年份：2019

卷号：40

期号：8

起止页码：65

中文期刊名：自动化仪表

外文期刊名：Process Automation Instrumentation

收录：CSTPCD

语种：中文

中文关键词：云技术;送餐机器人;深度强化学习;智能餐厅;控制系统

外文关键词：Cloud technology;Service robot;Deep reinforcement learning;Intelligent restaurant;Control system

摘要：随着人工智能的发展,机器人技术已成为研究热点。其中,用于餐厅服务的机器人更是近年来的关注热点。越来越多的餐厅选择使用机器人代替人工来完成相关任务。针对本地资源不足、餐厅服务机器人智能化程度不高等问题,设计餐厅服务机器人系统。基于云技术,分别从硬件和软件两方面设计餐厅智能服务机器人系统,包括送餐机器人、云端和客户端,以满足餐厅服务送餐需求。在机器人子系统中,使用深度强化学习,实现对未知环境路径规划的优化,提高机器人对餐厅复杂环境的适应性。仿真结果表明:所设计的机器人能很好地适应餐厅环境,及时避让障碍物,顺利完成送餐任务。与现有机器人相比,该系统克服了本地资源局限性和计算复杂性之间的矛盾,同时也为深度强化学习在机器人中的应用奠定了一定基础。
With the development of artificial intelligence,robot technologies have become a research hotspot.Among them,the robot used for restaurant service is a hot spot of concern in recent years.More and more restaurants choose to use robots instead of human resources to complete related tasks.To solve the problems of the limited local resources and the low level of intellectualization of restaurant service robots,the restaurant intelligent robot service system has been designed.Based on cloud technology,the restaurant intelligent robot service system from hardware to software,including the food delivery robot,cloud and client are designed to meet the demands for food delivery.In the robot subsystem,deep reinforcement learning is used to optimize the path planning in unknown environment and to improve the adaptability of the robot in the complex environment of the restaurant.The simulation results show that the designed robot can adapt to the restaurant environment,avoid obstacles in time,smoothly complete the food delivery tasks.Comparing with existing robot systems,the system proposed overcomes the contradiction between the limited local resources and the need of large amount of computational resources.At the same time,it lays a foundation for the application of deep reinforcement learning in robot systems.