基于二次规划的四足机器人机身轨迹优化研究

Quadratic programming based body trajectory optimization of quadruped robot

为使四足机器人在无地形感知运动过程中能表现出较稳定的运动性能,结合对角步态运动特点,提出了一种对角步态下零力矩点的约束方法。采用二次规划的优化方法分别在线优化出行走、对角步态下机器人稳定运动的机身轨迹。将优化出的两种步态下机身轨迹采用相同的映射关系映射到全身状态,结合规划控制框架,通过仿真验证两种步态下机器人的运动状态及轨迹跟随情况。仿真结果表明,机器人以行走步态运动时位置、速度跟踪效果及机身姿态稳定性相比对角步态较好,且两种步态下机器人都能跟随规划轨迹保持稳定运动。

In order to make the quadruped robot exhibit more stable motion performance during terrain-aware motion, this paper combines the characteristics of diagonal gait motion and proposes a constraint method with zero moment points under diagonal gait. A quadratic programming optimization method is used to optimize the robot body trajectory for stable motion under walking and diagonal gaits respectively online. The optimized trajectories are mapped to the whole-body state using the same mapping relationship, the motion state and trajectory following of the robot under the two gaits are verified by simulation in combination with the planning control framework. The simulation results show that the robot has better position and velocity tracking and body posture stability in walking gait than diagonal gait, and the robot can follow the planned trajectory and maintain stable motion under two gaits.