中枢模式发生器在六足机器人运动控制中的应用
发布时间：2018-11-05 12:15
【摘要】：近年来,机器人技术蓬勃发展,六足机器人作为机器人家族中极具典型的一员,已经在灾情救援、地表侦察、太空探测等诸多领域取得应用。同时运动控制作为机器人学的重要研究方向,得到了控制领域的广泛关注。随着生物学研究的逐渐深入,仿生控制方法在模拟生物运动节律的方向展现出相比传统控制方法更大的优越性,其中,以中枢模式发生器控制方法(Central Pattern Generator,CPG)最为典型。CPG控制方法能够在缺乏高层控制信号和外部反馈的情况下,产生稳定的节律输出信号,不需要对外部环境精确建模,控制方法简单有效,并可以很好地模仿生物运动步态。由此,本文将以CPG控制方法为核心,以六足机器人为控制对象,对其运动控制方法进行详尽的探讨。首先,论文介绍了关于步态周期、步长、占地比等基本概念,通过基于模型的控制方法,以机器人单腿为对象,进行几何分析和正逆运动学解算。在此基础上,对单腿足端轨迹进行多项式拟合,以此完成机器人的运动规划,保证运动准确性。同时,选择无刷直流电机作为机器人关节驱动装置,完成数学建模过程,并在关节空间内建立三闭环控制器,保证对高层的输出信号的精确跟踪。其次,通过采用中枢模式发生器中基于单一非线性振子的方法代替传统基于模型控制算法来进行关节轨迹生成。论文选取Hopf神经振荡器作为节律控制的基本单元,描述振荡器参数对输出波形的影响,利用串联惯性环节提供的相角滞后产生固定相位差的原始信号,并通过设置阈值线的方法确定单腿摆动相与支撑相的切换时刻,最后以调节采样频率以及补偿相位差的方式确定关节控制信号与原始相位差信号在离散时间内的一一对应关系,描述典型步态的规划发生过程。随后,针对基于单一振子的步态生成方法所产生的问题,论文提出了相应的方法予以改进。通过选取改进的Hopf振子实现摆动相与支撑相频率的单独可调,并通过利用CPG环状网络结构消除串联惯性环节带来的误差积累,整体上简化了控制过程,提高了仿生控制方法的运动控制效果。最后,论文以典型步态为例,通过六足机器人实验平台,对算法的正确性与可行性做出了验证。
[Abstract]:In recent years, robot technology has developed rapidly. As a typical member of robot family, hexapod robot has been applied in many fields such as disaster rescue, surface reconnaissance, space exploration and so on. At the same time, motion control, as an important research direction of robotics, has received extensive attention in the field of control. With the development of biological research, the bionic control method shows more advantages than the traditional control method in the direction of simulating biological motion rhythm. Among them, the central mode generator control method (Central Pattern Generator, CPG) is the most typical. CPG control method can produce stable rhythmic output signal without high level control signal and external feedback, and the control method is simple and effective without the need for accurate modeling of external environment. And can well imitate the biological movement gait. Therefore, the motion control method of hexapod robot is discussed in detail with CPG control method as the core and hexapod robot as the control object. Firstly, this paper introduces the basic concepts of gait period, step size and occupation ratio. The geometric analysis and forward and inverse kinematics are carried out through model-based control method. On the basis of this, polynomial fitting of the trajectory of one leg foot is carried out to complete the motion planning of the robot and ensure the accuracy of the motion. At the same time, the brushless DC motor is chosen as the robot joint driving device, and the mathematical modeling process is completed, and the three-loop controller is established in the joint space to ensure the accurate tracking of the output signals of the upper level. Secondly, the method based on single nonlinear oscillator in the central mode generator is used to generate the joint trajectory instead of the traditional model-based control algorithm. In this paper, Hopf neural oscillator is selected as the basic unit of rhythm control, and the influence of oscillator parameters on output waveform is described. The original signal with fixed phase difference is produced by the phase angle lag provided by series inertial link. The switching time between the swing phase and the support phase is determined by setting the threshold line. Finally, the one-to-one correspondence between the joint control signal and the original phase difference signal in discrete time is determined by adjusting the sampling frequency and compensating the phase difference. Describes the planning process of a typical gait. Then, in order to solve the problem of gait generation method based on single oscillator, the corresponding method is proposed to improve it. The frequency of swing phase and support phase can be adjusted separately by selecting improved Hopf oscillator, and the error accumulation caused by series inertial link is eliminated by using CPG ring network structure, which simplifies the control process as a whole. The motion control effect of the bionic control method is improved. Finally, taking typical gait as an example, the correctness and feasibility of the algorithm are verified by the experimental platform of hexapod robot.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP242
本文编号：2312053
[Abstract]:In recent years, robot technology has developed rapidly. As a typical member of robot family, hexapod robot has been applied in many fields such as disaster rescue, surface reconnaissance, space exploration and so on. At the same time, motion control, as an important research direction of robotics, has received extensive attention in the field of control. With the development of biological research, the bionic control method shows more advantages than the traditional control method in the direction of simulating biological motion rhythm. Among them, the central mode generator control method (Central Pattern Generator, CPG) is the most typical. CPG control method can produce stable rhythmic output signal without high level control signal and external feedback, and the control method is simple and effective without the need for accurate modeling of external environment. And can well imitate the biological movement gait. Therefore, the motion control method of hexapod robot is discussed in detail with CPG control method as the core and hexapod robot as the control object. Firstly, this paper introduces the basic concepts of gait period, step size and occupation ratio. The geometric analysis and forward and inverse kinematics are carried out through model-based control method. On the basis of this, polynomial fitting of the trajectory of one leg foot is carried out to complete the motion planning of the robot and ensure the accuracy of the motion. At the same time, the brushless DC motor is chosen as the robot joint driving device, and the mathematical modeling process is completed, and the three-loop controller is established in the joint space to ensure the accurate tracking of the output signals of the upper level. Secondly, the method based on single nonlinear oscillator in the central mode generator is used to generate the joint trajectory instead of the traditional model-based control algorithm. In this paper, Hopf neural oscillator is selected as the basic unit of rhythm control, and the influence of oscillator parameters on output waveform is described. The original signal with fixed phase difference is produced by the phase angle lag provided by series inertial link. The switching time between the swing phase and the support phase is determined by setting the threshold line. Finally, the one-to-one correspondence between the joint control signal and the original phase difference signal in discrete time is determined by adjusting the sampling frequency and compensating the phase difference. Describes the planning process of a typical gait. Then, in order to solve the problem of gait generation method based on single oscillator, the corresponding method is proposed to improve it. The frequency of swing phase and support phase can be adjusted separately by selecting improved Hopf oscillator, and the error accumulation caused by series inertial link is eliminated by using CPG ring network structure, which simplifies the control process as a whole. The motion control effect of the bionic control method is improved. Finally, taking typical gait as an example, the correctness and feasibility of the algorithm are verified by the experimental platform of hexapod robot.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP242
【参考文献】
相关期刊论文 前2条
1 许宏岩,付宜利,王树国,刘建国;仿生机器人的研究[J];机器人;2004年03期
2 张秀丽,郑浩峻,陈恳,段广洪;机器人仿生学研究综述[J];机器人;2002年02期
相关硕士学位论文 前1条
1 李文政;六足仿生机器人步态规划与控制系统研究[D];山东大学;2011年
，本文编号：2312053
本文链接：https://www.wllwen.com/kejilunwen/zidonghuakongzhilunwen/2312053.html
