QQ客服 官方微博 反馈留言

题名	针对未知连续曲面磨削机器人恒力跟踪控制算法研究
作者	刘哲
学位类别	硕士
答辩日期	2018-05-17
授予单位	中国科学院沈阳自动化研究所
授予地点	沈阳
导师	邹涛
关键词	机械臂 柔顺力控制 模型预测控制 遗传算法 阻抗控制
其他题名	Research on Constant Force Tracking Control Algorithm for Grinding Robot on Unknown Continuous Surface
学位专业	机械电子工程
中文摘要	本文针对未知连续曲面的磨削过程中恒力跟踪控制中具有超调、控制精度差和动态响应速度慢的问题进行了研究，主要的研究内容和成果如下：1.针对机械臂对未知连续曲面恒力跟踪具有较大超调的问题，提出了一种基于模型预测控制算法和曲面预测方程的曲面恒力跟踪方法。首先给出了受环境约束的机械臂离散、线性化的动力学模型；分析了机械臂末端和环境的力作用方式；设计了曲面预测方程，结合已走过的轨迹位置信息和力信息对未来的曲面位置进行估计；设计了基于模型预测控制算法的受约束机械臂轨迹跟踪力源控制方法；然后让机械臂根据该方法对估计的曲面位置进行力源跟踪。最后在仿真中控制机械臂追踪未知连续曲面，证明了机械臂对未知连续曲面作业时所提算法力控制的实时有效性，在保证控制精度的前提下，有效的抑制了超调问题，提高了系统的鲁棒性。2.受启发于弹性驱动器的工作原理，设计了一款主动弹性力控机构串联至机械臂末端用于提高机械臂曲面力跟踪的精度。首先给出了所设计末端的三维结构图和实体图；然后分析了机构的力源控制模型和不同刚度弹簧对控制结果的影响；最后通过仿真和实验，与传统的被动柔顺控制相对比，证明了该机构较传统被动柔顺控制方式更好的提高了机械臂末端接触力的恒力跟踪精度，减小了机械臂柔顺力控制对机械臂位置控制精度的要求。3.针对机械臂与环境接触时恒力跟踪时动态响应速度慢的问题，基于位置控制的阻抗控制方法和遗传算法，提出了基于实时优化遗传算法的阻抗控制方法。在研究过程中，首先对比了阻抗控制中三个阻抗参数对机械臂末端接触力的影响；然后根据提高机械臂恒力跟踪的响应速度和控制精度的性能指标，改进用于离线优化的遗传算法的交叉、变异和计算适应度值等操作算子的处理方式，实现了对阻抗控制方法中的参数进行实时优化；最后通过simulink和S-函数进行仿真。结果表明，和传统控制方法相比，该方法可以在保证控制精度的条件下，提高了机械臂与环境接触力的动态响应速度，并具有较小的或者没有超调量。
英文摘要	In this paper, the problem of constant force tracking in the grinding process of unknown continuous surfaces is studied. The main research contents and results are as follows. 1.Aiming at the problem of manipulator tracking the constant force of unknown continuous surface, a method of surface constant force tracking based on model predictive control algorithm and surface prediction equation is proposed. Firstly, the discrete and linear dynamic model of the environment constrained manipulator is given; the force acting modes of the manipulator arm and the environment are analyzed; the surface prediction equation is designed, which combines the track position information and force information that have already passed to the future. The position of the curved surface is estimated; a constrained manipulator trajectory tracking force source control method based on the model predictive control algorithm is designed; and then the manipulator arm is used to track the estimated surface position according to the method. Finally, in the simulation, the control manipulator tracks the unknown continuous surface, which proves the realtime effectiveness of the proposed algorithm's force control on unknown continuous surfaces. Under the premise of ensuring the control accuracy, it effectively suppresses the overshoot problem and improves the robustness of the system. 2. Inspired by the working principle of the elastic actuator (SEA), an active elastic force control mechanism was designed in series to the end of the arm. Firstly, the three dimensional structure diagram and the solid diagram of the designed terminal are given. Then the control model of the force source of the mechanism and the influence of different stiffness springs on the control results are analyzed. Finally, the simulation and experiment are compared with the traditional passive compliance control to prove Compared with the traditional passive compliance control method, this mechanism improves the constant force tracking accuracy of the manipulator arm contact force, and reduces the manipulator's position control precision requirements for the manipulator's compliance force control. 3.For the problem of constant force tracking when the manipulator is in contact with the environment, an impedance control method based on position control and a genetic algorithm are proposed, and an impedance control method based on a real- time optimization genetic algorithm is proposed. In the research process, the influence of the three impedance parameters in the impedance control on the end contact force of the manipulator arm is compared first; then, according to the performance index of improving the response speed and the control accuracy of the manipulator constant force tracking, the genetic algorithm used for offline optimization is improved. The crossover, variation, and calculation of fitness values, etc., are handled by the operator and realize the real time optimization of the three parameters in the impedance control method. Finally, simulations are performed using simulink and S-functions. The results show that compared with the traditional control method, this method can improve the speed of the contact force between the robot arm and the environment to converge to the desired contact force under the condition of ensuring the control accuracy, and has little or no overshoot.
语种	中文
产权排序	1
页码	71页
内容类型	学位论文
源URL	[http://ir.sia.cn/handle/173321/21748]
专题	沈阳自动化研究所_工业控制网络与系统研究室
推荐引用方式 GB/T 7714	刘哲. 针对未知连续曲面磨削机器人恒力跟踪控制算法研究[D]. 沈阳. 中国科学院沈阳自动化研究所. 2018.

个性服务

查看访问统计

相关权益政策

暂无数据

收藏/分享

除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。