基于CPG的四足机器人抗侧向冲击的动态稳定性研究

为解决四足机器人受到侧向冲击时的稳定性控制,提出了基于CPG和侧向踏步反射的控制方案:以Hopf振荡器构成的CPG网络为基础,通过为侧偏关节构造具有触发使能性质的振荡器,为四足机器人引入侧向踏步反射机制. 利用ZMP理论,引入倒立摆模型,从动力学角度预测四足机器人侧向踏步所需的步长与次数. 通过该反射,四足机器人在承受侧向冲击时产生的侧向加速度能够在较短时间内恢复正常,此后终止反射,配合正常直线行走控制方案,四足机器人就能够圆满实现在行走状态下受到侧向冲击后的稳定性控制,其抵御侧向冲击的能力显著提高. 利用Matlab与Adams联合仿真,该方案的可行性和有效性得到了可靠验证. 

CPG-Based Control Scheme for Quadruped Robot to Withstand the Lateral Impact

In the research of land mobile robots, the stability control of quadruped robots has gradually become the focus, also the difficulty. Aimed to keep stability to withstand lateral impact, a control scheme was proposed. A four triggered Hopf oscillator was introduced into the CPG network in the form of reflex to control the lateral movement of the robot. The algorithm, based on the ZMP theory and the inverted pendulum model, can predict the lateral step size and times to keep the stability. Worked in the normal walking control scheme, the lateral acceleration can quickly return to normal after the impact. And the ability to withstand lateral impact was significantly improved. The scheme was validated with the Matlab and Adams simulation.