一种新的双足机器人模型设计与相关研究

A New Kind of Bipedal Robot Model Design and Related Research

摘要:

针对双足机器人最简模型在行走过程中出现摆动腿足部擦地的问题,提出了一种通过摆动腿膝关节弯曲达到摆动腿缩短的新模型。当摆动腿开始摆动时,摆动腿膝关节弯曲锁定,摆动腿缩短；当摆动腿摆动到最大位置时,膝关节解锁,摆动腿伸直再锁定,此后摆动腿回摆,系统变为直腿模型。采用脚后跟冲击控制,在摆动腿落地前,拖后的支撑腿与地面接触处施加一指向髋关节的瞬时冲击力,冲击力可以减小摆动腿着地时能量的损耗,同时驱动被动机器人向前行走。设计了迭代学习控制算法,找到极限环与不动点,实现不同给定期望步长跟踪的冲击力的计算。仿真结果表明,迭代学习控制可以有效的实现不同期望步长的跟踪,可以很快的找到机器人系统的不动点,通过收敛的相平面,得到稳定的极限环,保证了机器人行走过程稳定。

Abstract:

When the simplest model of robot is in the process of walking, the feet of its swinging legs may brush the floor. In view of this problem, a model to shorten the swinging legs when the knees bend is proposed. When the swinging leg begins to swing, the knee of the swing leg bends and gets locked and the swing leg becomes shortened; when the swing leg swings to the largest position, the knee becomes unlocked and then locked after the swing leg unbends. In this condition, the swing leg swings back and the system becomes straight leg model. The heel impact control is adopted, so that before the swinging leg goes to the ground, an instantaneous wallop on to the hip can be applied at the place where the dragged support leg connect with the ground. The impact can reduce the energy loss of swinging leg landing on the ground, and can also drive the passive robot to walk forward. Iterative learning control algorithm is designed to find the limit cycle and the fixed point; through simulation, the impact forces under the tracking of different given expectation steps are calculated. The simulation results show that the iterative learning control can effectively implement the tracking of different expectation steps, quickly find the fixed points of robot system, and get a stable limit cycle through the convergence of the phase plane, to ensure the stable walking process of the robot.