面向核事故应急处置的遥操作系统设计与实验研究

Design and experimental study of teleoperation system for emergency response in nuclear accident

针对核事故现场的物理屏蔽严重、应急处置作业条件复杂、作业对象不明确等特点,研究并设计了一款基于动态自组网的机器人遥操作系统。在自组织网络实时Mesh保持良好通讯的情况下,利用叠加在真实回传视频中的虚拟模型,提出一种基于空间梯度的人工势场力觉引导算法。操作者可通过观察视觉交互系统中模型的变化和感受施加在自研主手的引导力,提高时延条件下的操作效率和临场感。实验表明,核辐射条件下遥操作系统130 m的通讯时延在30 ms以下,可满足视频传输的带宽要求。力觉引导下的物体避障抓取实验验证了算法的可行性和遥操作系统的工程实用性。因此,本文设计的基于动态自组网的机器人遥操作系统,可使操作者以更加直观自然的方式参与到遥操作系统中,有效提高操作员完成复杂遥操作任务的安全性和操作性能。

A robot teleoperation system based on dynamic self-organizing network is studied and designed to address the characteristics of severe physical shielding,complex emergency response conditions,and unclear operation objects at nuclear accident sites.A spatial gradient based artificial potential field force guidance algorithm is proposed using virtual models superimposed on real backhaul videos while maintaining good communication in a ad-hoc network’s real-time Mesh.Operators can improve operational efficiency and telepresence under time delay conditions by observing changes in the model in the visual interaction system and feeling the guiding force applied to the self-developed main hand.The experiment shows that under nuclear radiation conditions,the communication delay of the teleoperation system at 130 m is below 30 ms,which can meet the bandwidth requirements of video transmission.The experiment of obstacle avoidance and grasping guided by force guidance is verified the feasibility of the algorithm and the engineering practicality of the teleoperation system.Therefore,the robot teleoperation system based on dynamic self-organizing network designed in this article can enable operators to participate in the teleoperation system in a more intuitive and natural way,effectively improving the safety and operational performance of operators in completing complex teleoperation tasks.