6R工业机器人的运动轨迹规划及仿真研究

发布时间：2018-01-06 13:37

本文关键词：6R工业机器人的运动轨迹规划及仿真研究　出处：《中北大学》2017年硕士论文　论文类型：学位论文

【摘要】：随着机器人技术越来越成熟,企业制造商越来越广泛的使用机器人来完成工作任务。其中六自由度关节式机器人由于其操作空间大并且结构紧凑在制造业领域中广泛使用,它是生产制造中应用最广泛的机器人。由于关节式机器人的诸多优点,本文设计和分析了一个负载为6kg的关节式机器人。首先,对机器人常用的几种类型进行分析,最终选择关节式机器人进行设计,然后对其每个关节的结构进行设计,并建立solidworks三维模型图,使得机器人的整体结构更加直观;机器人要实现运动,就必须对其驱动系统进行设计,因此,根据经验及其他机器人样机,对工业机器人的电机和减速器进行选型;建立对工业机器人的连杆坐标系,得到其D-H参数,根据这些参数建立机器人的运动学方程,并运用Matlab Robotics toolbox工具箱对机器人的各个杆件进行建模,组成一个完整的机器人模型,对其进行正逆运动学和工作空间的仿真分析。其次,介绍了机器人的两种轨迹规划方法,这两种方法分别是关节空间轨迹规划和笛卡尔空间轨迹规划,然后对两者进行分析比较,综合两者的优缺点,最终本文选择关节空间进行轨迹规划;选取轨迹规划方法后结合matlab的强大功能对机器人的运动轨迹进行仿真,得到机器人末端的角位移、速度和加速度曲线,仿真得到的曲线光滑无突变,使得机器人可以按照预定的轨迹完成任务。最后,简单介绍RBF神经网络的结构,对RBF神经网络的matlab设计进行研究;由于RBF神经网络规模大、学习速度快、函数逼近拟合功能优的特点,将其运用到关节式机器人的轨迹规划中,结合matlab工具箱的强大功能,对机器人的轨迹进行仿真,得到机器人光滑的角位移曲线,神经网络得到的角位移曲线光滑,且与轨迹规划得到的曲线基本一致,说明了把RBF神经网络应用到工业机器人轨迹规划中是合理可行的。
[Abstract]:As robotics becomes more and more mature. Enterprise manufacturers are increasingly using robots to accomplish their tasks. Among them, six-degree-of-freedom joint robots are widely used in manufacturing due to their large operating space and compact structure. It is the most widely used robot in production and manufacture. Because of its many advantages, this paper designs and analyzes a robot with a load of 6kg. First of all. The common types of robot are analyzed. Finally, the joint robot is selected to design, and then the structure of each joint is designed, and the solidworks 3D model diagram is established. Make the whole structure of robot more intuitive; In order to realize the motion, the robot must design its drive system. Therefore, according to the experience and other robot prototype, the motor and reducer of the industrial robot are selected. The D-H parameters are obtained by establishing the coordinate system of the connecting rod of the industrial robot, and the kinematics equation of the robot is established according to these parameters. The Matlab Robotics toolbox toolbox is used to model each member of the robot to form a complete robot model. The forward and inverse kinematics and workspace simulation are analyzed. Secondly, two trajectory planning methods of robot are introduced, which are joint space trajectory planning and Descartes space trajectory planning. Then the two are analyzed and compared, and the advantages and disadvantages of the two are synthesised. Finally, the joint space is chosen for trajectory planning in this paper. After selecting the trajectory planning method and combining with the powerful function of matlab, the trajectory of the robot is simulated, and the angular displacement, velocity and acceleration curves at the end of the robot are obtained. Finally, the structure of RBF neural network is briefly introduced, and the matlab design of RBF neural network is studied. Due to the large scale, fast learning speed and excellent function of function approximation, RBF neural network is applied to the trajectory planning of joint robot and combined with the powerful function of matlab toolbox. The trajectory of the robot is simulated and the smooth angular displacement curve of the robot is obtained. The angular displacement curve obtained by the neural network is smooth and basically consistent with the curve obtained by the trajectory planning. It is reasonable and feasible to apply RBF neural network to industrial robot trajectory planning.
【学位授予单位】：中北大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP242

【参考文献】