双足机器人步态控制的C4.5决策树算法研究  

C4.5 Decision Tree Algorithm for Gait Control of Biped Robot

作　　者：郭丹丹 王宽田[2] GUO Dan-dan;WANG Kuan-tian(School of Software,Zhengzhou Institute of Industrial Application Technology,Henan Zhengzhou 450000,China;College of Ocean Engineering,Guilin University of Electronic Technology,Guangxi Guilin 541200,China)

机构地区：[1]郑州工业应用技术学院软件学院,河南郑州450000 [2]桂林电子科技大学海洋工程学院,广西桂林541200

出　　处：《计算机仿真》2024年第10期434-437,452,共5页Computer Simulation

摘　　要：为了提升双足机器人步态控制的稳定性,提高质心高度控制效果,提出一种面向双足机器人步态控制的C4.5决策树算法。通过模仿人类步行轨迹周期性变化,得出机器人步行的变化规律,运用正运动学计算出机器人关节运动角度以及实时状态信息,获得机器人质心的最优修正结果,修正质心轨迹,通过分析逆运动学求得双足机器人稳定运行条件。构建决策树模型,得出特征属性等数据集,基于信息熵以及决策树增益率构建出C4.5决策树,并对决策树实施剪枝处理,实现能够完成步态控制的C4.5决策树算法设计。实验结果表明,所提方法能够提升步态控制效果,保证机器人运动的稳定性。In order to improve the stability of gait control of bipedal robots,this paper presented a C4.5 decision tree algorithm for gait control of bipedal robots.At first,the changing rules of robot walking were obtained by imitating the periodic changes in human walking trajectory.Then,the positive kinematics was used to calculate the movement angle and real-time state information of the joint of the robot,and thus to obtain the optimal correction for correcting the center of mass trajectory.Through inverse kinematics analysis,the stable condition of the bipedal robot was obtained.A decision tree model was constructed to obtain feature attributes.Furthermore,a C4.5 decision tree was constructed based on information entropy and decision tree gain,and pruned.Finally,the design of the C4.5 decision tree algorithm for gait control was realized.Experimental results show that the proposed method can improve the effect of gait control and ensure the stability of robot motion.

关 键 词：双足机器人 步态控制 决策树 特征属性 轨迹规划 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]