助老助残机械手控制技术研究

发布时间：2018-06-17 12:42

本文选题：助老助残机械手 + 雅可比矩阵　；参考：《山东科技大学》2017年硕士论文

【摘要】：本论文主要对一种多自由度助老助残机械手控制系统的设计进行了研究。首先,根据助老助残机械手的结构特征,建立其D-H参数模型,求出正向运动学方程,并结合蒙特卡洛法对助老助残机械手的工作空间进行分析。推导了助老助残机械手的雅可比矩阵,为助老助残机械手的轨迹规划和速度控制的分析提供了理论依据。其次,由于助老助残机械手工作在非结构化环境中,针对该特点,采用了一种基于立体视觉的激光点识别定位的方法。立体视觉定位采用双目工业相机,通过阈值分割算法进行目标定位,不仅可以定位人脸嘴部的三维坐标,还可以通过激光引导定位空间内任一有效坐标。针对系统误差采用了基于BP神经网络的误差校正算法,并进行了实验仿真。然后,根据机器人学解耦理论,采用了一种基于广义回归神经网络的助老助残机械手逆向运动学解耦算法,将目标物体坐标反解出机械手各关节角度。最后,对助老助残机械手的位置和姿态进行了轨迹规划。通过在机械手小臂上安装一块MPU6050芯片采集出小臂的倾斜角度,根据该角度通过姿态补偿算法控制机械手小臂上两个电机的旋转使机械手末端手掌时刻保持平衡的姿态。由于在助老助残机械手的工作空间内可能存在障碍物,针对该问题,采用了一种基于人工势场法的机械手位置规划算法。
[Abstract]:In this paper, the design of a multi-degree of freedom assisted manipulator control system is studied. Firstly, according to the structural characteristics of the manipulator, the D-H parameter model is established, and the forward kinematics equation is obtained, and the workspace of the manipulator is analyzed with Monte Carlo method. The Jacobian matrix of the manipulator is derived, which provides a theoretical basis for the trajectory planning and speed control analysis of the manipulator. Secondly, because the assistive manipulator works in unstructured environment, a laser spot recognition and localization method based on stereo vision is adopted to solve this problem. A binocular industrial camera is used to locate the stereo vision, and the threshold segmentation algorithm is used to locate the target. It can not only locate the 3D coordinates of the face mouth, but also locate any effective coordinates in the space by laser guidance. The error correction algorithm based on BP neural network is adopted for the system error, and the experimental simulation is carried out. Then, according to the decoupling theory of robotics, an inverse kinematics decoupling algorithm based on generalized regression neural network is used to solve the different joint angles of the manipulator. Finally, the position and attitude of the manipulator are planned. By installing an MPU6050 chip on the forearm of the manipulator, the tilt angle of the forearm is collected, and the rotation of the two motors on the manipulator's forearm is controlled by the attitude compensation algorithm according to the angle. Because there may be obstacles in the workspace of the assisted manipulator, a position planning algorithm based on the artificial potential field method is adopted to solve the problem.
【学位授予单位】：山东科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP241

【参考文献】