基于5次B样条函数的果蔬采摘机器人轨迹规划的研究

发布时间：2018-07-01 08:14

本文选题：轨迹规划 + 人工势场　；参考：《浙江理工大学》2017年硕士论文

【摘要】：机器人各关节的轨迹规划是机器人控制的基础,机器人轨迹规划是根据机器人作业要求,在满足运动学约束情况下使机器人各个关节能够快速、平稳地依次通过各位置点,进而确保机器人末端执行器迅速平稳地到达预期位置进行作业,一方面可以提高机器人的工作效率,另一方面可以防止因为冲击过大降低机器人寿命。为了提高果蔬采摘机器人的工作效率和减少其在作业过程中所受的冲击,本文提出了基于时间-冲击两个运动性能指标,采用5次B样条插值函数进行轨迹规划,并利用GA算法对5次B样条插值轨迹进行优化,本文的主要内容如下:1、根据果蔬采摘机器人设计参数利用Creo2.0建立三维虚拟样机,根据DH坐标理论建立了机器人的数学模型,并根据DH参数求得机器人的运动方程,同时详细进行了机器人运动学反解过程的推算,验证了运动方程的正确性。此外根据DH参数利用MATLAB机器人工具箱建立了机器人的运动仿真模型,研究了机器人在工作空间运动情况。2、基于人工势场的思想将机器人的运动空间抽象成虚拟势场,根据合势场的数学模型,设计了机器人在动态环境作业时的避障算法,并利用MATLAB GUI设计了路径初始点、路径障碍及目标点的虚拟仿真软件,介绍了仿真软件的使用,通过构造单一障碍和多障碍虚拟环境进行仿真实验,实验结果表明机器人可以有效地避开工作环境中的障碍,证明了避障算法的正确性。3、研究了五次B样条的构造方法,并利用机器人工具箱函数对比分析了五次B样条和七次样条插值轨迹的异同点;根据机器人运动过程中所必须经过的位置点,建立关于时间-冲击的目标函数,利用GA算法获得优化后五次B样条插值轨迹,并利用MATLAB仿真出机器人在优化之后运动轨迹,与前面非优化轨迹相比,优化之后的运动轨迹缩短了6.62s,并且各个关节在运动过程中所受的冲击也得到明显减小。
[Abstract]:The trajectory planning of robot joints is the basis of robot control. According to the requirements of robot operation, robot trajectory planning enables each joint of robot to pass through each position point quickly and smoothly under the condition of satisfying kinematics constraints. In order to ensure that the robot end actuators can reach the desired position quickly and smoothly, on the one hand, it can improve the robot's working efficiency, on the other hand, it can prevent the robot from reducing the robot's life because the impact is too big. In order to improve the working efficiency of the fruit and vegetable picking robot and reduce the impact of the robot during the operation, this paper presents two motion performance indexes based on time-impact, and adopts the B-spline interpolation function of 5 times to carry out trajectory planning. The interpolation locus of B-spline is optimized by GA algorithm. The main contents of this paper are as follows: 1. According to the design parameters of fruit and vegetable picking robot, the 3D virtual prototype is built by Creo2.0, and the mathematical model of robot is established according to DH coordinate theory. The kinematic equation of the robot is obtained according to the DH parameters, and the inverse kinematic solution of the robot is calculated in detail, which verifies the correctness of the equation. In addition, according to DH parameters, the robot motion simulation model is established by using MATLAB robot toolbox, and the robot motion in workspace is studied. Based on the idea of artificial potential field, the robot motion space is abstracted into virtual potential field. According to the mathematical model of potential field, the obstacle avoidance algorithm of robot in dynamic environment is designed, and the virtual simulation software of path initial point, path obstacle and target point is designed by MATLAB GUI, and the use of simulation software is introduced. By constructing a single obstacle and multi-obstacle virtual environment, the simulation results show that the robot can effectively avoid the obstacles in the working environment, and prove the correctness of the obstacle avoidance algorithm. The method of constructing the quintic B-spline is studied. The similarities and differences of the interpolating trajectories of the quintic B-spline and the cubic spline are analyzed by using the robot toolbox function, and the objective function about time-impact is established according to the position points that the robot must pass through in the course of motion. The fifth B-spline interpolation trajectory is obtained by GA algorithm, and the motion trajectory of the robot after optimization is simulated by MATLAB, which is compared with the previous non-optimal trajectory. After optimization, the motion trajectory is shortened by 6.62 s, and the impact of each joint during the motion process is obviously reduced.
【学位授予单位】：浙江理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP242

【参考文献】