摘要
针对四足机器人动态转向失稳与非结构化地形适应性不足的问题,提出四足机器人转弯控制及非结构化地形自适应优化算法。通过融合Denavit-Hartenberg运动学约束与多模态感知,优化质心-足端力矩分配与地形刚度的动态匹配。实验结果表明:算法轨迹跟踪最大均方根偏差较最优对比方法降低64%(标准差0.04,精度达标率86.7%);能量效率系数较对比方法的极值降低58%,稳定区间(18.0~25.0 J/m)覆盖率达92.3%,能耗波动率下降42%;足端滑移率均值仅0.33。非结构化地形验证证实其能同时保持控制精度达标率大于85%及能耗波动率小于15%。所提算法提高了四足机器人的运动控制精度,能够为四足机器人的全地形运动优化提供全新方案。
Aiming at the dynamic steering instability and insufficient adaptation to unstructured terrain in quadruped robots,this study proposes an optimization algorithm for turning control and terrain adaptation.By integrating Denavit-Hartenberg kinematic constraints with multimodal sensing,the algorithm optimizes centroid-foot-end moment distribution for dynamic terrain stiffness matching.Experimental results demonstrate that the algorithm achieves a maximum root mean square error(RMSE)reduction of 64%compared to the optimal reference method(standard deviation 0.04,accuracy compliance rate 86.7%).The energy efficiency coefficient decreases by 58%relative to the reference method's peak value,with a stable range coverage rate of 92.3%(18.0~25.0 J/m)and a 42%reduction in energy consumption fluctuation rate.The mean foot slip rate is only 0.33.Validation in unstructured terrain confirms simultaneous maintenance of>85%control accuracy and<15%energy consumption fluctuation.This work significantly enhances quadruped robot motion control precision and delivers a novel solution for all-terrain motion optimization.
作者
李信言
安娜
王婷
Li Xinyan;An Na;Wang Ting(Personnel Division,Xi'an Aeronautical Polytechnic Institute,Shaanxi Xi'an,710089,China)
出处
《机械设计与制造工程》
2026年第2期75-80,共6页
Machine Design and Manufacturing Engineering
基金
2024年度陕西省教育厅科研计划(24JK0509)。
