Design and Tension Modeling of a Novel Cable-Driven Rigid Snake-Like Manipulator

作     者：Xu, Dawei Li, En Liang, Zize Gao, Zishu 

作者机构：Chinese Acad Sci Inst Automat State Key Lab Management & Control Complex Syst Beijing 100190 Peoples R China Univ Chinese Acad Sci Beijing 100049 Peoples R China 

出 版 物：《JOURNAL OF INTELLIGENT & ROBOTIC SYSTEMS》 (智能和机器人系统杂志)

年 卷 期：2020年第99卷第2期

页      面：211-228页

核心收录：

学科分类：08[工学] 0811[工学-控制科学与工程] 0812[工学-计算机科学与技术（可授工学、理学学位）] 

基　　金：National Key R&D Program of China [2018YFB1307400] National Natural Science Foundation of China [61873267, U1713224] 

主　　题：Snake-like manipulator Cable-driven Kinematics Tension model Neural network Reinforcement learning 

摘      要：In this study, a cable-driven snake-like manipulator with high load capacity and end-positioning accuracy is designed for applications in complex and narrow environments. The Hooke joint-like two degree-of-freedom joint design and the full actuation mode enhanced the rigidity of the robot. The modular link design increased the local flexibility of the robot. Because the cable tension cannot be ignored under high load on the basis of the kinematics model, a cable tension model is established based on rigid body static equilibrium to describe the relationship between posture and cable tension. This provided a foundation for follow-up studies on obstacle avoidance path planning with optimized tension. At the same time, in order to improve the response speed of the motor position controller to the tension change, this study introduces both the tension as the reference model input and the system state variable into the adaptive control method based on model identification and reinforcement learning. The kinematics model and the cable tension model were validated by experiments on the prototype. The practical results of the two adaptive control methods were compared and the result shows that the method based on model identification has a better effect.