六轴机器人控制器底层软件设计及其算法仿真研究

发布时间：2018-05-26 00:25

本文选题：运动控制器 + 机器人运动学　；参考：《深圳大学》2017年硕士论文

【摘要】：机器人控制器是机器人的控制核心,也是衡量机器人运动性能水平的重要标志,其底层运动学算法设计的好坏直接影响着机器人的运动控制性能。本课题以六轴工业机器人为研究对象,以伺服驱动技术为基础,从运动学算法层次,对工业机器人运动控制器底层的开发设计做出了系统性介绍。机器人运动学研究。通过D-H建模法得到机器人目标位置姿态矩阵;采用几何解析和坐标变换的方法分析出机器人的8组位置运动学逆解以及对应的8种机器人结构状态,以8种结构状态为基础分析出了机器人的3种奇异位型,并对奇异位型提出了具体的处理措施。通过雅克比矩阵分析了如何通过速度运动学实现其轨迹控制,将速度运动学的控制效果和位置运动学的控制效果进行了比较,分析二者的优劣。运动学轨迹插补算法。基于螺旋理论提出了姿态的运动规划策略;使用多项式样条插值完成了直角坐标和关节坐标的规划轨迹的设计;通过数学几何分析和坐标变换完成机械手的空间直线路径规划、空间圆弧路径规划及空间柱状螺旋线路径规划的设计;基于梯形加减速的速度规划方式完成了对多段线连续轨迹前瞻控制的设计。运动控制器底层软件设计。对控制器底层架构做出总体分析,分别从FPGA中底层逻辑模块设计和DSP中的底层软件设计两方面对其开发设计过程作出了系统性的全面介绍。算法仿真及控制器的离线和在线测试。通过MATLAB工具箱完成了对机器人的建模,并以此为基础完成了各种规划轨迹的仿真研究和运动学算法的验证。通过opengl软件库完成了对3ds格式的机器人3D文件的读入,并在visual studio中开发出了机器人运动控制的仿真软件,该仿真软件可以通过TCP/IP协议与运动控制器进行以太网通信,通过将运动控制算法写入到运动控制器,由运动控制器向仿真软件发送机器人关节的控制数据来驱动图形仿真机器人的关节运动,并画出运动轨迹来验证运动控制器内部运动控制算法的可行性。将控制器连接到驱动机器人关节的伺服电机上,在Debug模式下通过预先给定的运动规划轨迹测试机器人本体的实际运动情况。
[Abstract]:Robot controller is not only the core of robot control, but also an important symbol to measure the level of robot motion performance. The design of bottom kinematics algorithm directly affects the performance of robot motion control. Taking six-axis industrial robot as research object and servo drive technology as the foundation, this paper systematically introduces the development and design of motion controller of industrial robot from the level of kinematics algorithm. Kinematics of robots. The pose matrix of robot's target position is obtained by D-H modeling method, eight sets of inverse kinematics solutions of robot's position and corresponding eight kinds of robot's structural states are analyzed by geometric analysis and coordinate transformation. On the basis of eight structural states, three singular bit types of the robot are analyzed, and the specific treatment measures for the singular bit types are put forward. By means of Jacobian matrix, how to realize the trajectory control by velocity kinematics is analyzed. The control effect of velocity kinematics and the control effect of position kinematics are compared, and the advantages and disadvantages of the two methods are analyzed. Kinematics trajectory interpolation algorithm. Based on the helical theory, the motion planning strategy of attitude is proposed, the planning trajectory of rectangular coordinate and joint coordinate is designed by polynomial spline interpolation, and the space linear path planning of manipulator is accomplished by mathematical geometric analysis and coordinate transformation. The design of space arc path planning and spatial cylindrical spiral path planning, and the speed planning method based on trapezoidal acceleration and deceleration completed the design of multi-segment continuous trajectory forward control. Motion controller bottom software design. This paper makes a general analysis on the basic architecture of the controller and makes a systematic and comprehensive introduction to the development and design process of the controller from the following two aspects: the design of the underlying logic module in FPGA and the design of the underlying software in the DSP. Algorithm simulation and off-line and online testing of the controller. The modeling of the robot is completed by MATLAB toolbox, and based on this, the simulation research of various planning trajectories and the verification of kinematics algorithms are completed. The 3D file of robot in 3ds format is read in by opengl software library, and the simulation software of robot motion control is developed in visual studio. The simulation software can communicate with the motion controller via TCP/IP protocol. The motion control algorithm is written to the motion controller, and the motion controller sends the control data of the robot joint to the simulation software to drive the motion of the robot. The motion trajectory is drawn to verify the feasibility of the motion control algorithm in the motion controller. The controller is connected to the servo motor that drives the robot joints, and the actual motion of the robot body is tested under the Debug mode through the pre-given trajectory of motion planning.
【学位授予单位】：深圳大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP242

【参考文献】