摘要
随着人类活动范围的不断扩大,焊接技术被广泛地应用到了空间、水下以及核环境的一些装配、维修作业中。在这些极限环境中,用机器人代替人来执行焊接任务是必要的。由于目前相关技术的发展还不能满足机器人完全自主焊接的需要,有人参与的遥控焊接技术受到了重视。
在智能化的遥控焊接系统中,需要对焊接质量进行动态的实时监控、对机器人的运行轨迹进行实时跟踪、对整个焊接过程进行远程监控,必然会引入或涉及到多个具有一定独立性、自主性并具有特定功能的资源。这些物理资源决定了机器人遥控焊接系统存在着内在的,包括时间、空间以及功能上的分散性。这就使得利用多智能体技术实现机器人遥控焊接系统中的协调控制具有其可能性、必然性并且具有一定的实践意义。
本文从实际应用角度出发,介绍了实现遥控焊接的基本硬件结构,分析了遥控焊接的生产过程及控制方式,并应用多智能体技术对机器人遥控焊接系统进行了智能体结构建模。
对机器人遥控焊接多智能体间的通信及协作机制进行了分析,采用点对点的消息通信方式,以KQML语言作为消息封装格式,KIF知识交换格式作为通信内容的知识描述;在分析了基于协商的协作方法、合同网协商协议的基础上,提出了按需设岗,竞争上岗的协作模型。
构建了一套机器人遥控焊接多智能体系统,集成了机器人控制系统、遥操作系统、激光跟踪系统、力觉传感系统以及焊缝熔透控制系统。利用多智能体分散式控制模式,通过各个资源智能体实现对资源的管理、控制。
最后,对激光跟踪传感智能体、力觉传感智能体、焊接机器人智能体以及操作者协同工作的遥示教过程进行了实验验证,实验表明基于多智能体技术的机器人遥控焊接系统具有较高的效率、精度和稳定性。
With the expanding of human activities, welding technology is widely applied to some assembly and maintenance work underwater, in space and nuclear environment. It is necessary to use robot to execute weld task instead of human in the extreme environment. Now relevant technologies do not yet permit a totally autonomous remote welding operation, tele-robotic welding with the help of operators have been paid more attentions.
In an intelligent tele- robotic welding system, it is greatly necessary to monitor the quality of seam and track the path of robot real-time, and monitor the processing of welding remotely. So physical device resources with certain independence and automation are imported. Because of these physical resources tele-robotic welding system possesses some inherent distributed characteristics. Thus it is possible and necessary to control tele- robotic welding system with the distributed multi-agent cooperation technology.
In the viewpoint of application, the basic hardware configuration of tele-robotic welding system has been firstly introduced. And then on the basis of analyzing the production process and control mode of tele-robotic welding, the multi-agent model of tele-robotic welding system has been presented.
The communication and cooperation mechanism of tele-robotic welding system were studied. Message transmitting in point to point, message encapsulating in KQML, message contents described in KIF were selected for information exchanges. By analyzing negotiation based coordination methods and Contract Net Protocol, the cooperation model of“station based on requirement, mount guard through competition”is presented.
We have constructed a tele-robotic welding multi-agent system, which contains welding robot, tele-operation system, laser scanning welding seam tracking sensor, force sensor and seam molten pool monitoring system. Based on the distributed control characteristics of multi-agent system, all subsystems are monitored, managed and coordinated through corresponding resource agent.
At last tele-teaching experiments by the cooperating of laser scanning welding seam tracking agent, force agent, welding robot agent and operator were fulfilled.
引文
1 王立春. 多agent协商与通信研究. 南京大学博士学位论文. 2001:2-10
2 H. V. D. Parunak. Distributed AI and manufacturing control: Some issues and insights. Proceedings of 1st European Workshop on Modeling an Autonomous agent in a multi-agent world. 1989:30-37
3 廖强,周凯,张伯鹏. 基于现场总线的多Agent作业车间动态调度问题的研究.中国机械工程. 2000,11(7):757-759
4 Cohen G. Concurrent system to resolve real-time conflicts in multi-robot systems. Engineering Application Artificial Intelligence. 1995, 8(2):169-175
5 陈仁际,吴镇炜,王韬等. 分布式多机器人装配系统任务合作规划算法研究.中国机械工程. 2000,1(4):418-421
6 N. R. Jennings et al. Agent Technology: Foundations, Application, and Market, Springer-Verlag Heidelberg. 1997:1-5
7 W. B. Parsley. Equipment used for Examining, Dismantling, and Reassembly of a HFIR Fuel Element. Proceedings of 17th Conference on Remote Systems Technology. 1969:209-214
8 J. E. Agapakis. Vision-Aided Remote Robotic Welding. Ph.D. dissertation, MIT. 1-50
9 J. J. Conrath. Remotely Controlled Repair of Piping at Douglas Point. International Conference on Robotics and Remote Handling in the Nuclear Industry, Toronto, Canada. 1984:112-121
10 H. Trevor, L. Trevor. ARM and Rovsim: extending our reach. Industrial Robot. 1999, 26(3):202-208
11 张惠滨. 主从遥控弧焊机器人实验系统建立和操作特性研究. 吉林工业大学焊接专业博士学位论文. 1994:12-40
12 M. Hou. On Tele-operation of an Arc Welding Robotic System. Proceedings- IEEE International Conference on Robotics and Automation. 1996, 2:1275-1280
13 吕伟新. 遥控弧焊运动控制新方法研究. 哈尔滨工业大学博士学位论文. 1997:1-80
14 朴永杰. 焊接柔性制造单元多智能体协调控制的研究. 上海交通大学博士学位论文. 2004:2-30
15 梁泉. 多智能体分散式协调控制的研究. 上海交通大学博士学位论文. 1996:1-50
16 陈白皋. 基于 Agent 的智能控制系统的分析、设计与实现. 中国科学技术大学硕士学位论文. 2000:1-2
17 王汝传,徐小龙,黄海平等. 智能 Agent 及其在信息网络中的应用. 北京邮电大学出版社, 2006:22-27,31-41,44-45
18 B. Dunin-Keplicz. Compositional Formal Specification of Multi-agent systems. 1994:1-5
19 C. Hewit. Viewing Control Structures as Patterns of Passing Messages. Artificial Intelligence. McGraw-Hill. New York, USA. 1977,8(3):323-364
20 高锷. 基于多Agent的车间制造系统控制结构及控制技术研究. 合肥工业大学博士学位论文. 2003:3-73
21 D. M. Lane, A. G. Mcfadzean. Distributed problem solving and real-time mechanisms in robot architectures. Engineering Application Intelligence, 1994, 7(2):105-117
22 C. Ramos. Architecture and a negotiation protocol for the dynamic scheduling of manufacturing systems. 1994:3161-3166
23 Y. C. P. Jeff, Tenenbaum. An intelligent agent framework for enterprise intelligation. IEEE Trans on Systems, Man and Cybernetics, 1991, 21(6): 1391-1408
24 F. Sprumont, J. P. Muller. AMACOIA: A multi-agent system for designing flexible assembly lines. Applied Artificial Intelligence. 1997, 11(6): 573-589
25 B. C. Draa, P. Millot. A framework for cooperative work: An approach based on the intentionality. Artificial Intelligence in Engineering. 1990, 5(4): 199-205
26 S. J.Russell. Provably bounded optimal agents. Proc of the 13 th Int Joint Conf on Artificial Intelligence. 1993: 40-48
27 H. Wang, C. Wang. APACS: A multi-agent system with respository support. Knowledge-based Systems. 1996, 9(3): 329-337
28 J. Ferber. Multi-Agent Systems. The Second Edition. Addition Wesley Longman Inc., 1999:36-37
29 N. R. Jennings, K. Sycara, M. J. Wooldridge. A roadmap of agent research and development. Journal of Autonomous Agents and Multi-agent Systems. 1998, 22(4): 275-306
30 Bernardo Huberman, H. Clearwater. A multi-agent system for controlling building environment. Proceedings of the first international conference on multi-agent systems. 1995:171-176
31 史忠植等. 智能主体研究现状与发展趋势. 计算机世界. 1998,4:1
32 史忠植. 智能主体及其应用. 科学出版社, 2000:30-37
33 张洁,高亮,李培根. 多 Agent 技术在先进制造中的应用. 科学出版社,2004:13-20
34 陈兰芳. 多智能体技术及其在生产系统协调控制中的应用. 河北工业大学硕士学位论文. 2005:8-14
35 胡文斌. 基于多 Agent 的分布式智能群决策支持系统关键技术研究. 武汉理工大学博士学文论文. 2004:71-89
36 T. Finin, D. McKay, R. McEntire. KQML as an Agent Communication Language. In Proc. Third Int. Conf. On Inf. and Knowledge Management, ACM Press, 1994:1-3
37 张丽霞. 多 Agent 技术及其在供应链管理中的应用研究. 天津大学硕士学位论文. 2004:14-18
38 陈红普. 基于多 Agent 的水下机器人通信机制研究. 哈尔滨工程大学硕士学位论文. 2005:21-30
39 陈武华,王立春,李红兵等. 扩展 KQML 以实现合同网. 计算机工程与应用. 2000,(4):106-109
40 杨涛. 基于多智能体的移动机器人控制与协调研究.华中科技大学硕士学文论文. 2005:10-15
41 R. G. Smith. The contract net protocol: High-level communication and control in a distributed problem solver. IEEE Transactions on Computers. 1980, 29(12): 1104-1113
42 周伟. 基于多智能体协作的虚拟企业任务调度和冲突消解研究. 中南大学博士学位论文. 2004:45-57
43 张连新,高洪明,吴林等. 基于 PMAC 的开放式弧焊机器人控制系统. 焊接学报. 2003,24(6):85-88