【摘要】 针对现阶段机器人远程反向操控能力的不足，借助协作机器人的灵活性与数字孪生的虚实融合能力，本研究开发了一种实时反向操控的数字孪生系统，并对虚实同步动作的位置偏差与同步时延展开研究。位置偏差方面，本研究借助物理场景特征点对机器人基坐标系位姿进行标定，并据此对孪生场景进行修正；同步时延方面，本研究通过建立引擎外部通信代码代替在内部调用接口的方式，降低了因接口调用阻塞造成的动作滞后时间。最后，通过数字孪生物理场景进行验证，测试结果表明该系统能够实现低位置偏差、低时延的物料抓取与拧紧枪取钉等反向操控动作，系统虚实同步效果良好，该方法也为高温、辐射、人机协作与远程医疗等场景提供了孪生交互方案。更多还原

【Abstract】 In response to the current lack of remote reverse control capabilities for robots, this study leveraged the flexibility of collaborative robots and the virtual-real integration of digital twins to develop a real-time reverse control digital twin system. The study investigated both position deviation and synchronization delay in virtual-real synchronous actions. Regarding positional deviation, physical scene feature points were used to calibrate the position of the robot’s base frame, with corresponding adjustments made to the twin scene. For synchronization latency, the study mitigated action delays caused by interface call blocking by implementing external communication code for the engine, as opposed to internal interface calls. Finally, experimental validation in digital twin physical scenes demonstrated that the system achieves low position deviation and minimal latency in material grasping and reverse control tasks, such as tightening and nail picking. The system exhibits strong virtual-real synchronization, offering a viable twin interaction solution for applications in high-temperature environments, radiation-prone areas, human-robot collaboration and remote medical care.更多还原