双足直立行走机器人步态规划及其控制研究

发布时间：2019-02-21 21:35

【摘要】：本课题的主要任务是研究仿人形双足直立行走机器人的控制,核心是双足直立行走机器人步态规划算法和步态控制算法。在国内外机器人研究的理论基础上进行总结创新,本文主要工作和成果如下:通过引入李群和旋量理论,建立了机器人腿部机构的指数积正运动学模型,结合正运动学模型的指数积公式,提出了一种分步求解关节角度的方法,其中膝关节和踝关节处的三个角度直接通过几何法求解,髋关节处的三个角度用矩阵法。利用同样的方法,建立机器人手臂以及头部并联机构的运动学模型。通过研究三维倒立摆的数学模型,得到了三维倒立摆运动规律简洁的参数化表达方法,即双曲函数表示法。同时研究了机器人前进方向,基于双曲正弦的运动规律和左右方向基于双曲余弦的运动规律,提出了一种步态落点预测控制算法。针对机器人起步、停步以及迈步的轨迹规划研究,提出了限幅最速微分跟踪器和最速运动规划算法,利用这两种算法进行轨迹规划,轨迹平滑全程无速度突变、无冲击,同时能够到达任意目标位置并获得目标速度。为有效控制腿部振动,采用了一种带模型参数的自抗扰算法,改进自抗扰算法中的扩展观测器,有效解决利用不同类型的物理反馈信号,观测角度、角速度信号的难题,实际实验中大大提高了系统的稳定性和鲁棒性。利用机器人姿态角度偏差,通过PI控制机器人姿态角度,实现机器人在倾斜面上平衡控制。为了提高姿态角度的解算精度,提出了一种新的互补滤波融合方法,不仅利用重力加速方向作为参考,而且充分利用了重力加速度大小,减少加速度对角度解算的影响。最后,完成机器人步态规划以及机器人控制实物实验,编写了Labview和Matlab的调试上位机用于实物实验之前的程序调试。通过实物实验验证了运动学模型、步态规划算法以及机器人控制算法的有效性和正确性。
[Abstract]:The main task of this paper is to study the control of humanoid biped upright walking robot, the core of which is gait planning algorithm and gait control algorithm. The main work and achievements of this paper are as follows: by introducing Li Qun and spinor theory, the exponential product forward kinematics model of robot leg mechanism is established. Based on the exponential product formula of the positive kinematics model, a step by step method is proposed to solve the joint angle, in which the three angles of the knee joint and the ankle joint are directly solved by the geometric method, and the three angles of the hip joint are solved by the matrix method. Using the same method, the kinematics model of robot arm and head parallel mechanism is established. By studying the mathematical model of 3D inverted pendulum, a simple parameterized expression method of motion law of 3D inverted pendulum is obtained, that is hyperbolic function representation method. At the same time, the forward direction of the robot is studied. Based on the motion law of hyperbolic sine and the motion of left and right directions based on hyperbolic cosine, a gait drop predictive control algorithm is proposed. Aiming at the research of trajectory planning for starting, stopping and stepping of robot, this paper presents a limiting maximum differential tracker and a maximum speed motion planning algorithm. These two algorithms are used for trajectory planning. At the same time can reach any target position and obtain target speed. In order to control leg vibration effectively, an active disturbance rejection algorithm with model parameters is adopted to improve the extended observer in the active disturbance rejection algorithm, which effectively solves the problem of using different types of physical feedback signals, observation angles and angular velocity signals. The stability and robustness of the system are greatly improved in practical experiments. The attitude angle of the robot is controlled by PI to realize the balance control of the robot on the inclined plane by using the attitude angle deviation of the robot. In order to improve the accuracy of attitude angle calculation, a new complementary filtering fusion method is proposed, which not only uses the acceleration direction of gravity as a reference, but also makes full use of the magnitude of gravity acceleration to reduce the influence of acceleration on angle calculation. Finally, the robot gait planning and the robot control physical experiment are completed, and the Labview and Matlab debugging host computer is programmed to debug the program before the physical experiment. The validity and correctness of kinematics model gait programming algorithm and robot control algorithm are verified by physical experiments.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP242

【相似文献】