基于二维力反馈的机械手夹持器设计与研究
本文选题:夹持器 + 八角环二维力传感器 ; 参考:《江南大学》2017年硕士论文
【摘要】:随着智能制造飞速发展,工业机器人作为一种先进的制造装备扮演着不可或缺的角色。由于需要完成各种各样的任务,机器人需要各种不同的末端执行器来完成物料的抓取。在食品包装、农业采摘等行业,对于柔软或酥脆物品的抓取一直是机器人技术应用的难点。如何使机械手像人手一样,能依靠复杂的感知系统,检测物体信息和夹持力信号,准确地识别物体的软硬程度、重量、表面情况等信息,精确控制抓取力的大小,是目前国内外智能机器人研究的重点之一。本文针对食品机器人应用需求,设计了一款小型八角环二维力传感器,用来感知抓取过程中夹持力和切向力信息。在此基础上设计了机器人灵巧末端夹持器,实现对柔软或酥脆物品的稳定抓取。首先,对夹持器抓取物体过程进行了受力分析,确定切向力与夹持力的检测需求。设计了能够检测二维力的八角环传感器,利用ANSYS仿真软件对传感器进行了有限元和应变节点分析,对其结构参数进行优化设计。设计制作了八角环传感器实验样机和检测装置,通过传感器静态标定实验得到耦合矩阵,再反解得到解耦矩阵,能够通过补偿减少因为机械加工不准和应变片位置粘贴不到位而产生的误差。通过简单的夹持试验证明,该八角环传感器满足对物体夹持力与切向力的检测要求。设计制作了夹持器样机和系统实验平台,包含夹持器、控制系统硬件和上位机软件等。控制系统硬件由工控机、运动控制卡、数据采集卡、交流伺服系统、八角环二维力传感器及信号放大器、线性直流电源等组成。上位机软件用LabVIEW开发,具有实验数据采集、分析以及和机器人控制系统通讯接口功能。利用实验平台对夹持力的控制方法进行了研究。夹持器在抓取物体时会产生一定的形变,因此采用位置和力的串级控制进行夹持力控制。在位置控制阶段采用变速控制,以减小夹持器由张开状态到接触物体过渡时间。针对不同刚性物体,采用不同的抓取速度,并且采用基于力闭环变速控制的算法,即将目标力和实际力的力偏差转化为位置控制系统的速度,当力偏差减小,夹持器二指闭合速度自动减速直至停止,从而减小力的超调量和过大的振荡。实验结果表明,夹持力控制系统稳定,能快速的达到参考力,满足最终的抓取要求。研究了基于二维力反馈的动态抓取策略。在抓取过程中,通过八角环二维力传感器检测反馈夹持力与切向力的二维力信息,采用基于二维力反馈的力比例控制,使切向力与夹持力的比值保持在一定的阈值以内,从而使物体和夹持器手指之间不发生滑动,保证夹持器手指不脱离物体表面,并且保护物体避免因夹持力过大而被损坏。实验证明了该方案能成功抓取物体且避免损害。
[Abstract]:With the rapid development of intelligent manufacturing, industrial robots play an indispensable role as an advanced manufacturing equipment. Because of all kinds of tasks, robots need different end actuators to capture materials. In food packaging, agricultural picking and other industries, grasping soft or crisp items has always been a difficult point in the application of robot technology. How to make the manipulator, like a hand, be able to rely on complex sensing systems to detect object information and clamping force signals, accurately identify the soft and hard degree of the object, weight, surface conditions, and so on, and accurately control the size of the grasping force. It is one of the emphases of intelligent robot research at home and abroad. In order to meet the requirements of food robot application, a small octagonal two dimensional force sensor is designed to sense the information of clamping force and tangential force during grasping. On this basis, the robot dexterous end gripper is designed to achieve a stable grasp of soft or crisp objects. Firstly, the mechanical analysis of gripper grabbing process is carried out to determine the measurement requirements of tangential force and clamping force. The octagonal ring sensor which can detect two-dimensional force is designed. The finite element analysis and strain node analysis of the sensor are carried out by using ANSYS simulation software, and the structural parameters are optimized. The experimental prototype of octagonal ring sensor and the detecting device are designed and manufactured. The coupling matrix is obtained by the static calibration experiment of the sensor, and then the decoupling matrix is obtained by inverse solution. It can reduce the error caused by machining inaccuracy and strain gauge position misbonding by compensating. It is proved by simple clamping test that the octagonal ring sensor can meet the requirements of measuring the clamping force and tangential force. The prototype of the gripper and the experimental platform of the system are designed and manufactured, including the gripper, the hardware of the control system and the software of the upper computer. The hardware of the control system consists of industrial control computer, motion control card, data acquisition card, AC servo system, octagonal two dimensional force sensor and signal amplifier, linear DC power supply and so on. The upper computer software is developed by LabVIEW and has the functions of data acquisition, analysis and communication with robot control system. The control method of clamping force is studied by using the experimental platform. The gripper will produce certain deformation when grabbing the object, so the position and force cascade control is used to control the clamping force. In the position control stage, variable speed control is used to reduce the transition time from open state to contact object. Aiming at different rigid objects, different grasping speed is adopted, and the algorithm based on force closed loop variable speed control is adopted. The force deviation of target force and actual force is transformed into the speed of position control system, and the force deviation decreases when the force deviation is reduced. The two-point closing speed of the gripper automatically slows down until it stops, thus reducing the overshoot of force and excessive oscillation. The experimental results show that the clamping force control system is stable, can quickly reach the reference force, and meet the final grasp requirements. The dynamic capture strategy based on two-dimensional force feedback is studied. In the process of grasping, the 2-D force information of the feedback clamping force and tangential force is detected by the octagonal ring two-dimensional force sensor, and the force proportional control based on the two-dimensional force feedback is adopted to keep the ratio of the tangential force to the gripping force within a certain threshold. Therefore, there is no sliding between the object and the gripper finger, which ensures that the gripper finger does not break off the surface of the object and protects the object from being damaged by excessive clamping force. Experiments show that the scheme can grasp objects successfully and avoid damage.
【学位授予单位】:江南大学
【学位级别】:硕士
【学位授予年份】:2017
【分类号】:TP242.2
【参考文献】
相关期刊论文 前10条
1 桂树国;常礼;马成学;曹会彬;孙玉香;高理富;;六维力传感器设计与实验数据分析[J];仪表技术与传感器;2014年11期
2 胡志勇;张学炜;张伟;王琳;;西瓜采摘末端执行器夹持力精确控制[J];农业工程学报;2014年17期
3 周俊;杨肖蓉;朱树平;;基于自适应神经模糊网络的果蔬抓取力控制[J];农业机械学报;2014年07期
4 周俊;朱树平;;农业机器人果蔬抓取中滑觉检测研究[J];农业机械学报;2013年02期
5 王学林;姬长英;周俊;姜莉;顾宝兴;;基于灰色预测控制的果蔬抓取系统设计与试验[J];农业工程学报;2010年03期
6 孔民秀;游玮;杜志江;孙立宁;;基于速度闭环的自适应力位控制算法[J];哈尔滨工业大学学报;2010年03期
7 张俊强;周军;丁希仑;;一种自适应二指夹持器的设计及应用[J];机电工程;2008年05期
8 李秦川;胡挺;武传宇;胡旭东;应义斌;;果蔬采摘机器人末端执行器研究综述[J];农业机械学报;2008年03期
9 韩京清;;自抗扰控制技术[J];前沿科学;2007年01期
10 唐润宏;陈文楷;余跃庆;;松下交流伺服系统与PMAC控制卡在机器人中的应用[J];仪器仪表用户;2006年05期
相关博士学位论文 前2条
1 张庭;仿人型假手指尖三维力触觉传感器及动态抓取研究[D];哈尔滨工业大学;2014年
2 王学林;农业机器人末端执行器抓持力控制研究[D];南京农业大学;2009年
相关硕士学位论文 前3条
1 李雪峰;基于滑动检测的假肢手反射控制研究[D];中南大学;2014年
2 朱树平;基于滑觉检测的农业机器人果蔬抓取研究[D];南京农业大学;2012年
3 田强;基于仿人智能控制夹持器力控制系统研究[D];重庆大学;2012年
,本文编号:1924062
本文链接:https://www.wllwen.com/kejilunwen/zidonghuakongzhilunwen/1924062.html