基于CPG的四足机器人稳定性优化及偏航补偿

CPG-based stability optimization and yaw compensation for quadruped robots

针对处于步态摆动相时四足机器人重心偏移而引起得的机体偏航,以及严重情况下俯仰角过大引起的机身不稳定的问题,提出一种基于智能优化和自适应补偿结合的解决方案。首先,基于Hopf振荡器与相应足间运动关节相位耦合并引入运动控制参数,建立中枢模式发生器(central pattern generator, CPG)的网络运动模型;然后,提出一种改进的自适应调整权重的麻雀搜索算法(Adaptive Weight SSA),并用该算法通过迭代寻找输出步态信号最优参数组合,降低重心偏移产生的机体俯仰角与偏航值。最后,在改进麻雀搜索算法优化后四足机器人稳定性提升,使用最小二乘法得出偏航函数模型预测实时的偏航增长量,并使用偏航补偿函数控制CPG模型所需关节转动补偿偏航量。本文在Python环境下的Pybullet虚拟仿真平台上进行仿真实验。经四足机器人运动仿真结果证明了提出的方法的有效性和合理性。四足机器人通过偏航补偿控制函数解决了偏航问题,所提算法寻优提高了行走稳定性。

In order to solve the problems of yawing caused by a shift in the centre of gravity of the quadruped robot during the gait swing phase, and in severe cases the instability of the body caused by excessive pitch angle. A solution based on a combination of intelligent optimization and adaptive compensation is proposed. Firstly, a central pattern generator(CPG) based on Hopf model oscillator is coupled with the corresponding motion joints and motion control parameters are introduced to establish the CPG network motion model;After that, an improved Adaptive Weight Adjustment Sparrow Search Algorithm(Adaptive weight SSA) is proposed and used to find the optimal combination of parameters for the output gait signal through iteration to reduce the pitch angle and yaw value of the body generated by the shift in the center of gravity. Finally, the stability of the quadruped robot is improved after the optimization of the improved sparrow search algorithm, the yaw function model is derived using least square method to predict the yaw growth in real time, and the yaw compensation function is used to control the joint rotation required by the CPG model to compensate for the yaw. In this paper, simulation experiments are conducted on the Pybullet virtual simulation platform in Python environment. The effectiveness and rationality of the proposed method are proved by the simulation results of the quadruped robot motion. The research of quadruped robot in the paper solves the yaw problem by the yaw compensation control function, and the proposed algorithm seeks to improve the walking stability.