面向松耦合服务的机器人集成框架及应用研究  

Research on Framework and Application of Integrated Robot System Based on Loose Coupled Service

作　　者：张国亮[1] 王钰 王展妮[1] 林昌龙[1] ZHANG Guo-liang, WANG Yu, WANG Zhan-ni, LIN Chang-long(Computer Science and Technology Institute, Huaqiao University, Xiamen 361021, Chin)

机构地区：[1]华侨大学计算机科学与技术学院,福建厦门361021

出　　处：《小型微型计算机系统》2018年第4期651-656,共6页Journal of Chinese Computer Systems

基　　金：国家自然科学基金项目(61403150)资助;福建省自然科学基金项目(2016J01302)资助

摘　　要：针对机器人软件控制系统对扩展性和重用性的应用需求,将分布式组件服务、多代理协调与层次化软件设计思想结合,构建了基于松耦合服务的机器人集成智能框架.在服务层,将机器人与网络服务统一定义为分布式组件服务,忽略了操作系统及编程语言带来的系统移植性问题,解除了上层应用与底层硬件的强耦合关系.在协调层,将AI领域中多Agent结构与本体论系统描述结合,研究了基于MASO结构的多代理协调组件管理方法.基于此框架,针对APF路径规划中典型的局部极小问题,研究了基于变换行走策略的改进势场规划方法.仿真结果证实了所提框架及规划方法的可行性.To meet the application need of augmentability and reusability of robotic software control system,distributed component service,multi-agent coordination and hierarchical software design are combined to build an intelligent integrating robotic framework based on loose coupled service in this paper.In service layer,robot and web service are all served as distributed component service,which not only overcome portability problems caused by OS and programming language,but also remove the strong coupled relationship between upper application and lower hardware.In coordination layer,by integrating multi-agent structure and Ontology system description in artificial intelligent field,multi agent coordination for managing component is investigated based on MASO.Within this framework,traditional APF path planning is improved by switching strategies to solve local minima problems.The simulation shows the proposed framework and planning method is feasible.

关 键 词：多代理系统 分布式软件服务 松耦合服务 人工势场法 局部极小 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]