计算机工程与应用 ›› 2017, Vol. 53 ›› Issue (8): 113-119.DOI: 10.3778/j.issn.1002-8331.1509-0096

• 模式识别与人工智能 • 上一篇    下一篇

基于自然语言的分拣机器人解析器技术研究

熊志恒,闵华松   

  1. 武汉科技大学 冶金自动化与检测技术教育部工程研究中心,武汉 430081
  • 出版日期:2017-04-15 发布日期:2017-04-28

Research on sorting robot parser based on natural language understanding

XIONG Zhiheng, MIN Huasong   

  1. Engineering Research Center for Metallurgical Automation and Measurement Technology, Ministry of Education, Wuhan University of Science and Technology, Wuhan 430081, China
  • Online:2017-04-15 Published:2017-04-28

摘要: 工业机器人通常采用特定的机器人语言进行示教编程与控制,对于操作人员需要具有较高专业与技能要求,并且示教周期长导致工作效率降低。为了提高工业机器人使用效率与易用性,提出一种基于受限自然语言解析器的设计方法。该系统通过对受限自然语言进行词法解析、语法解析、语义解析,得到所需求的工作意图,然后与实时生成的三维空间语义地图进行匹配,结合机械臂轨迹规划,生成能够完成工作任务的机器人作业程序,并完成了机器人作业程序的解析与实际机械臂的控制。通过实验证明设计的基于受限自然语言处理的分拣机器人解析器能够正确解析自然语言命令,实现对机械臂的控制。

关键词: 受限自然语言, 机器人语言, 解析器, 轨迹规划

Abstract: Industrial robots usually adopt specific robot language to teach programming and controlling, which needs high professional skills for operators. In addition, longer period of teaching programming can reduce the efficiency. In order to raise the efficiency and the applicability of industrial robots, this essay presents a kind of design approach based on constrained natural language parser. By lexical parsing, grammar parsing and semantic parsing to constrained natural language, this system can get the needed working intent, afterwards, it matches with real-time producing three-dimensional semantic map. Combing manipulator trajectory planning, it eventually generates robot operating program that can finish the targets. Besides, it finishes parsing the robot operating program and controlling the actual manipulator. By doing experiment, it appears that the robot parser handled based on limited natural language in this essay can correctly parse natural language commands, finally can achieve the controlling to the manipulator.

Key words: constrained natural language, robot language, parser, trajectory planning