软体机械臂水下自适应鲁棒视觉伺服  

Adaptive Robust Visual Servoing Control of a Soft Manipulator in Underwater Environment

作　　者：徐璠 王贺升[1] XU Fan;WANG He-Sheng(Department of Automation,School of Electronics Information and Electrical Engineering,Shanghai Jiao Tong University,Shanghai 200240)

机构地区：[1]上海交通大学电子信息与电气工程学院自动化系,上海200240

出　　处：《自动化学报》2023年第4期744-753,共10页Acta Automatica Sinica

基　　金：国家自然科学基金(62073222,61722309)资助。

摘　　要：水下仿生软体机器人在水底环境勘测,水下生物观测等方面具有极高的应用价值.为进一步提升仿章鱼臂软体机器人在特殊水下环境中控制效果,提出一种自适应鲁棒视觉伺服控制方法,实现其在干扰无标定环境中的高精度镇定控制.基于水底动力学模型,设计保证动力学稳定的控制器;针对柔性材料离线标定过程繁琐、成本高,提出材料参数自适应估计算法;针对水下特殊工作条件,设计自适应鲁棒视觉伺服控制器,实现折射效应的在线补偿,并通过自适应未知环境干扰上界,避免先验环境信息的求解.所提算法在软体机器人样机中验证其镇定控制性能,为仿生软体机器人的实际应用提供理论基础.Underwater bioinspired soft robots enjoy high applicability in underwater exploring,biology observing,etc.This paper proposes an adaptive robust controller to improve the performance of the octopus-inspired soft robot in underwater environment and to realize accurate positioning control in the uncalibrated environment with disturbances.The paper designed a dynamically stable controller based on the underwater dynamic model.Considering the high-costing and tedious offline identification of soft material parameters,the paper proposed an adaptive method to estimate the unknowns.Considering the special underwater working condition,the paper proposed an adaptive robust visual servo controller,which can online compensate for refraction effect and avoid solving prior environment information by estimating unknown upper bound of environment disturbances.The proposed algorithm was experimentally validated in a soft robot prototype,serving as a theoretical basis of applications of bioinspired soft robots.

关 键 词：软体机器人 自适应鲁棒控制 无标定视觉伺服 水下视觉伺服 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置] TP391.41[自动化与计算机技术—控制科学与工程]