机器人关节伺服系统的复合控制

发布时间：2018-03-01 20:03

本文关键词： 机器人控制永磁同步电机模糊神经网络控制滑模控制位置控制与轨迹规划　出处：《青岛大学》2017年硕士论文　论文类型：学位论文

【摘要】：机器人是一种由各种电机驱动的装置,驱动电机的工作性能影响着机器人的工作性能。然而,传统研究中很多机器人控制方法的研究只对机器人动力学进行研究,没有考虑电机的因素。本文将关节机器人模型的控制方法研究和驱动电机运动控制的研究结合。永磁同步电机因结构简单、运行可靠度高和维护方便等诸多优点,在本文机器人关节伺服系统中得到应用。机器人具有较强的非线性和不确定性,给机器人的控制研究带来很多困难。为了解决这个问题,本文以二自由度关节机器人为研究对象,将永磁同步电机作为机器人关节的驱动电机。本文研究了一种PD加重力补偿的控制方法,该方法在PD控制器的基础上引进重力补偿项,能够加快系统的响应速度,减少达到稳定状态的时间。电机控制器中选用滑模控制方法,这种方法对电机具有良好的控制效果。仿真结果表明,基于PD重力补偿的机器人控制系统具有良好的响应特性和跟踪能力。针对PD加重力补偿控制中运动响应和跟踪能力的一些不足,提高机器人的控制效果,研究了一种基于模糊神经网络的机器人控制方法。模糊神经网络是一种兼具模糊控制和神经网络控制各自优点的控制方法。由于其良好的控制效果和学习能力,可替代PD加重力补偿控制,能够解决机器人控制中的非线性和不确定的问题。在使用模糊神经网络控制时,先对机器人动力学进行分离,将其中的不确定项分离,然后利用模糊神经网络的方法进行控制。同时为了减小滑模控制造成的抖动,对上一部分提出的滑模控制方法进行了一些改进。仿真结果表明,对比PD加重力补偿控制,系统的跟踪和响应能力得到显著提高。机器人在运动时容易受到外界的各种干扰,一个具有良好性能机器人伺服系统不仅要求系统能够准确迅速的跟踪输入值,而且要求系统能有较好的抗扰动能力。为了解决这个问题,本文研究了带有负载转矩观测器的机器人关节控制方法。一般情况下可以将外部扰动视为控制系统负载转矩的变化。所以将在模糊神经网络控制器的基础上增加负载转矩观测,提高系统的抗干扰能力。仿真结果表明,该控制方法在受到转矩扰动时,受到的影响较小,具有良好的抗干扰能力。综上,为了解决机器人的非线性和不确定问题,提高系统的快速响应跟踪能力和抗干扰能力,本文研究了几种不同的控制器。通过对不同控制器的控制效果进行对比分析,对系统进行相应的改进。本文最后研究的带有负载转矩观测器的基于模糊神经网络和滑模控制的机器人伺服控制系统,不仅具有良好的快速响应和跟踪能力,而且具有良好的抗干扰能力达到了设计要求。
[Abstract]:Robot is a kind of device driven by a variety of motors. The working performance of the driving motor affects the performance of the robot. However, many traditional research methods of robot control only study the dynamics of the robot. This paper combines the control method of joint robot model with the motion control of driving motor. The permanent magnet synchronous motor has many advantages, such as simple structure, high operation reliability and convenient maintenance, etc. The robot has strong nonlinearity and uncertainty, which brings a lot of difficulties to the research of robot control. In order to solve this problem, In this paper, the permanent magnet synchronous motor (PMSM) is used as the driving motor of the robot joints, and a control method of PD weighting force compensation is studied, which introduces the gravity compensation term based on the PD controller. It can speed up the response speed of the system and reduce the time to reach the stable state. The sliding mode control method is used in the motor controller, which has a good control effect on the motor. The simulation results show that, The robot control system based on PD gravity compensation has good response characteristics and tracking ability. In this paper, a robot control method based on fuzzy neural network is studied. Fuzzy neural network is a control method with the advantages of both fuzzy control and neural network control. Instead of PD weighting force compensation control, it can solve the nonlinear and uncertain problems in robot control. In the use of fuzzy neural network control, the dynamics of robot is separated first, and the uncertainty is separated. Then the fuzzy neural network is used to control it. In order to reduce the jitter caused by sliding mode control, some improvements are made to the sliding mode control method proposed in the previous part. The simulation results show that the PD weighting compensation control is compared with PD weighting force compensation control. The tracking and response ability of the system has been greatly improved. The robot is vulnerable to various kinds of interference when it moves. A robot servo system with good performance not only requires the system to track the input value accurately and rapidly. In order to solve this problem, the system is required to have better anti-disturbance ability. In this paper, a robot joint control method with a load torque observer is studied. In general, the external disturbance can be regarded as the change of the load torque of the control system. Therefore, the load torque observation will be added to the fuzzy neural network controller. The simulation results show that the control method is less affected by torque disturbance and has good anti-jamming ability. In order to solve the nonlinear and uncertain problems of the robot, the simulation results show that the proposed control method has good anti-interference ability. In order to improve the ability of fast response tracking and anti-jamming, several kinds of controllers are studied in this paper, and the control effects of different controllers are compared and analyzed. Finally, the robot servo control system with load torque observer based on fuzzy neural network and sliding mode control is studied. And has good anti-interference ability to meet the design requirements.
【学位授予单位】：青岛大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP242

【参考文献】