多自由度机械手臂的控制技术研究

发布时间：2018-06-23 15:01

本文选题：机械手臂 + 路径规划　；参考：《安徽工业大学》2017年硕士论文

【摘要】：随着中国制造2025规划的实施和我国制造业的崛起,工业机器人和各类机械手臂的应用越来越普及,对机械手臂控制技术的的研究越来越受到重视。机械手臂控制关键技术就是机械手臂的路径规划方法和轨迹跟踪控制算法。论文运用数学方法搭建了多自由度机械手臂的运动学模型,分析研究了运动学正、逆问题,侧重对运动学中逆问题进行了分析探讨;机械手臂动力学问题一样包含正、逆问题,本论文从动力学数学建模和最优控制等几个方面的问题进行了简约表述。发现路径规划中碰撞点的预知和路径点的设定都与机械手臂末尾控制器的工作空间有关。非奇异路径规划是指机械手臂在完成指定任务时规划出一条有效避开障碍物的路径,对此论文提出了一种改进型算法,此算法有效的结合牛顿迭代法和关节坐标规划法,用此种改进的算法可使得机械手臂(包括奇异位形的情况)规划出合理的路径,即在执行任务时能够顺利的通过工作空间中奇异区域,而不会造成错位、突然变速等异常情况改进的算法可以很好的解决路径规划中奇异性的问题。在运动轨迹跟踪方面,运用了模糊神经网络算法来的实现轨迹跟踪,此算法巧妙地结合了模糊控制强大的推理能力和神经网络控制的参数自学习能力,它不需要知道被控对象精确的数学模型,可以直接利用此算法对网络参数进行自学习并做出相应的修正,所以本论文把模糊神经网络算法运用在多自由度机械手臂的关节力矩控制,通过此算法对网络参数的自学习来调整机械手臂关节力矩的控制,从而实现对机械手臂的轨迹跟踪控制。最后经过试验仿真分析,把模糊神经网络控制算法运用在多自由机械手臂的控制上可以有效的解决轨迹跟踪问题,使整个控制系统有较强的适应性和稳定性。
[Abstract]:With the implementation of the 2025 Plan of manufacture in China and the rise of China's manufacturing industry, the application of industrial robots and various kinds of mechanical arms is becoming more and more popular, and more attention has been paid to the research on the control technology of mechanical arms. The key technology of robot arm control is path planning method and trajectory tracking control algorithm. In this paper, the kinematics model of multi-degree-of-freedom manipulator is built, and the forward and inverse kinematics problems are analyzed and studied. In this paper, the dynamic mathematical modeling and optimal control are briefly described. It is found that the prediction of collision point and the setting of path point in path planning are related to the workspace of the controller at the end of the manipulator. The nonsingular path planning means that the robot arm can plan a path to avoid obstacles effectively when completing the assigned task. An improved algorithm is proposed in this paper, which effectively combines Newton iterative method and joint coordinate planning method. With this improved algorithm, the robot arm (including the case of singular configuration) can be planned out a reasonable path, that is, the task can be carried out smoothly through the singular region of the workspace without causing dislocation. The improved algorithm can solve the singularity problem in path planning. In the aspect of motion trajectory tracking, the fuzzy neural network algorithm is used to realize trajectory tracking. This algorithm skillfully combines the powerful reasoning ability of fuzzy control with the parameter self-learning ability of neural network control. It does not need to know the exact mathematical model of the controlled object. It can directly use this algorithm to self-learn the network parameters and make corresponding corrections. So this paper applies the fuzzy neural network algorithm to the joint torque control of multi-degree-of-freedom mechanical arm, and adjusts the joint torque control of mechanical arm by self-learning of network parameters. Thus, the trajectory tracking control of the robot arm is realized. Finally, the fuzzy neural network control algorithm is applied to the control of multi-free manipulator, which can effectively solve the trajectory tracking problem, and make the whole control system have strong adaptability and stability.
【学位授予单位】：安徽工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP241;TP183

【参考文献】