摘要
针对四足仿生机器狗运动避障控制精度的要求,结合四足仿生机器狗运动学模型,提出了一种基于模型预测控制的运动控制方法,并探讨在不同控制参数和不同工况下模糊预测控制器的规避效果。研究结果表明,在不同控制参数和参考速度下,控制器能让四足仿生机器狗按照给定的避障轨迹进行跟踪,且跟踪轨迹误差控制在10 cm以内,偏航角误差稳定在2°以内。通过以上研究可知:基于模糊预测控制的控制器具有良好的稳定性,能较好地实现四足仿生机器狗的步态控制;同时,不同误差权重系数Q和权重系数R参数设置对机器狗的运动控制效果也不同。Q和R越小,其稳定性越强,但响应速度越慢。
According to the requirements of motion obstacle avoidance control accuracy of quadruped bionic robot dog,combined with the kinematics model of quadruped bionic robot dog,a motion control method based on model predictive control is proposed,and the avoidance effect of fuzzy predictive control and working conditions is discussed.The results show that under different parameters and reference speeds,the controller can make the quadruped bionic robot dog track according to the given obstacle avoidance trajectory,and the tracking trajectory error is controlled within 10 cm,and the yaw angle error is stable within 2°.The above research shows that the controller based on fuzzy predictive control has good stability and can better realize gait control of quadruped bionic robot dog.At the same time,different parameter settings of error weight coefficient Q and weight coefficient R have different effects on the motion control of robot dog.The smaller Q and R are,but the slower their response speed is.
作者
周永强
李佳
崔晔
孙强
ZHOU Yongqiang;LI Jia;CUI Ye;SUN Qiang(State Grid Baoding Power Supply Company,Baoding 071051,China)
出处
《自动化仪表》
CAS
2022年第8期79-83,共5页
Process Automation Instrumentation
基金
国网河北省电力有限公司科技基金资助项目(kj2020-014)。
