多关节机器人运动学与轨迹规划及仿真研究

发布时间：2018-01-01 11:07

本文关键词：多关节机器人运动学与轨迹规划及仿真研究　出处：《吉林大学》2016年硕士论文　论文类型：学位论文

【摘要】：随着工业机器人广泛地在焊接、搬运、码垛、喷涂等领域大显身手,工业机器人的应用领域变得更加宽广。在应用过程中,对机器人空间位姿具有很严格的要求,但却不能通过示教的方式来解决这一问题。因此,必须对机器人逆运动学问题进行求解,并且在逆运动学解的基础上进行多解优化和完成轨迹规划,最终替代示教器的功能,提高机器人的智能水平,实现机器人离线编程。针对不满足PIEPER准则的6R机器人的逆运动学解算法,存在计算过程特别复杂,或者不能求出逆运动学方程。所以本文提出一种基于多种群连续遗传算法的机器人逆运动学算法。通过多种群算法,避免算法陷入局部最小值,提高算法收敛精度;通过连续遗传算法,加快算法的收敛速度。通过对全部机器人逆运动学解进行一定规则的优化,得到一个最优的工程逆运动学解;再通过对机器人关节角度进行规划,得到一个最优的机器人连续运动的角度时间关系。以松下TA1400机器人作为求解的实例表明,本文提出的算法计算逆运动学解不受机器人几何结构的限制,并且末端位姿误差可以精确到小数点后2位,可在高精度轨迹跟踪问题中应用。本文主要研究内容如下:1)阐述了国内外对机器人逆运动学求解、逆解的多解优化和关节空间轨迹规划以及三维建模仿真的研究现状,介绍了课题的研究意义、背景。研究了空间位姿的表示方法以及多关节机器人的运动学问题。对机器人的工作空间进行分析,为后续的一般几何结构机器人逆运动学解的求取打下了良好的基础。2)对遗传算法原理进行简单介绍。针对该算法的收敛速度慢和收敛精度不高等问题,提出多种群连续遗传算法。对一个一般几何结构的6自由度机器人的逆运动学解问题,同时利用基本遗传算法和改进后的算法进行求取,得到改进后的算法的收敛速度和收敛精度具有较大的提高,该算法可以在工程实际中应用。同时,对后面的多解优化以及关节角度规划打下了基础。3)对机器人逆运动学解的多解问题进行研究。通过对机器人划线过程的分析,对逆解进行多解优化,得到一个唯一的工程解。对多解优化后的结果在关节角度内进行直线插补、抛物线插补、多项式插补和样条曲线插补,分析各种插值结果的优劣性。依据各种插补的结果,同时参考笛卡尔空间轨迹的长度,为轨迹跟踪打下良好的基础。4)验证各种插值结果对轨迹跟踪的精度影响。对于各种轨迹规划进行MATLAB编程,对各种关节轨迹规划后的角度、速度、加速度以及加加速度进行仿真,检验它们是否连续,最后得到合理的规划方法。编写机器人三维虚拟仿真软件,对机器人逆运动学解以及多解优化后的结果进行仿真研究。本文通过对不满足PIEPER结构机器人逆运动学解、机器人逆解的多解优化和关节角度空间插值等方面进行相关研究,得出一个适用于机器人划线的算法,并用MATLAB的机器人工具箱对上述问题进行验证。最后,本文利用CATIA建立各个连杆几何模型,应用VC++的MFC模块和Open GL功能开发了机器人三维仿真软件,成功实现了对机器人逆运动学、轨迹规划等问题的仿真和验证。
[Abstract]:With the industrial robot widely in welding, handling, palletizing, spraying and other fields display skills to the full field of application of industrial robots, has become more broad. In the application process, has very strict requirements for the space robot pose, but not by teaching the way to solve this problem. Therefore, the inverse kinematics problem of the robot must to solve multi solution optimization and complete trajectory planning based on inverse kinematics solution, eventually replace the teaching device function, improve the intelligent level of the robot, the robot off-line programming. The inverse kinematics algorithm does not meet the criteria of PIEPER 6R robot, existing calculation process is very complicated, or not inverse the kinematics equation. So this paper proposes an inverse kinematics algorithm for multi population genetic algorithm based on continuous robot. Through a variety of swarm algorithm, the algorithm avoids local most Small values, improve the convergence accuracy; through continuous genetic algorithm to accelerate the convergence speed. The inverse kinematics of robot optimization rules for all solutions, get an optimal engineering inverse kinematics; through the planning of the robot joint angle, get an optimal robot continuous movement angle of time. By TA1400 show that solving the robot as Example, this paper presents an algorithm for computing the inverse solution is not affected by the geometry of the robot kinematics constraints, and pose errors of the end effector can be accurate to 2 decimal places, can be used in high precision trajectory tracking problem. The main research contents of this paper are as follows: 1) of the robot at home and abroad are described. Inverse kinematics, inverse optimization and trajectory planning in joint space and the research situation of 3D modeling and Simulation of multi solutions, introduces the research significance, the background of subject research. The spatial position of the representation and kinematics of multi joint robot. To analyze the working space of the robot, as a general geometric structure of robot subsequent inverse kinematics solution for foundation.2 good) makes a brief introduction to the principle of genetic algorithm. The convergence speed of the algorithm is slow convergence and low precision the problem, put forward multi population genetic algorithm. For the inverse problem of kinematics of a general geometric structure of the robot with 6 degrees of freedom at the same time, using the basic genetic algorithm and the improved algorithm is obtained, the convergence speed and precision of the improved algorithm has greatly improved, the algorithm can be used in engineering practice at the same time, the solution of optimization and joint angle planning behind the multi lay the foundation of.3) of robot inverse kinematics solution of the multi solution problem is studied. By scribing on Robot The analysis on the inverse solution for multi solution optimization, get a unique solution. The multi solution optimized results are linear interpolation, in the joint angle parabolic interpolation, polynomial interpolation and spline interpolation, advantages and disadvantages analysis of various interpolation results. According to the interpolation results, at the same time in Cartesian space the length of the reference trajectory for trajectory tracking, lay the foundation for.4 good effect) verification of various interpolation results on the accuracy of trajectory tracking. MATLAB programming for a variety of trajectory planning, speed of joint trajectory planning after the angle, acceleration and jerk simulation, to check whether they are continuous, finally, reasonable planning methods preparation of 3D virtual simulation software. The robot, inverse kinematics and the multi solution after optimization of robot simulation. Based on the PIEPER structure to meet the robot inverse The robot kinematics, inverse kinematics and multi solution optimization of joint angle spatial interpolation and other aspects of research, get a suitable algorithm to robot marking, and the above problems are verified by MATLAB robot toolbox. Finally, CATIA established each link geometric model used in this article, the application of VC++ MFC module and Open GL function development the robot 3D simulation software, the successful implementation of the robot inverse kinematics, simulation and verification of trajectory planning problems.

【学位授予单位】：吉林大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TP242

【相似文献】