基于姿态信息的双足移动机器人越障步态控制
Gait Control of Biped Mobile Robot Based on Attitude Information
刘晓铭 1华定忠 2张静 1黄国方 1钟亮民1
作者信息
- 1. 南瑞集团(国网电力科学研究院)有限公司,江苏南京 211106;国电南瑞科技股份有限公司,江苏 南京 211106
- 2. 南瑞集团(国网电力科学研究院)有限公司,江苏南京 211106
- 折叠
摘要
在机器人运动控制中,姿态信息获取是不可缺少的步骤,尤其是在机器人越障步态控制中,因此提出基于姿态信息的双足移动机器人越障步态控制方法.在双腿摆动和地面碰撞2种情境下,创建双足移动机器人动力学模型,获得机器人行走变化规律.利用AdaBoost算法分类机器人姿态属性,通过关联性指标与骨骼密度实现机器人姿态信息感知,明确操作者越障意图.使用摄像头得到障碍物图像数据,采用神经网络的前馈补偿能力进行机器人越障步态控制.实验结果表明,所提方法越障步态控制精准度高、稳定性强,可提高双足移动机器人的移动能力.
Abstract
In robot motion control,attitude information acquisition is an indispensable step,especially in robot gait control over obstacles.Therefore,a gait control method for biped mobile robot over obstacles based on attitude information is proposed.Under the two situations of legs swinging and ground collision,the dynamic model of biped mobile robot is created,and the walking change law of robot is obtained.The AdaBoost algorithm is used to classify the robot's attitude attributes,and the robot's attitude information perception is realized through the association index and bone densityto clarify the operator's intention to surmount obstacles.The camera is used to get the obstacle image data,and the feedforward compensation capability of neural network is used to control the robot's gait over obstacles.The experimental results show that the proposed gait control method has high accuracy and can improve the mobility of biped mobile robots.
关键词
姿态信息/双足机器人/越障步态控制/动力学模型/神经网络Key words
attitude information/biped robot/gait control for obstacle crossing/dynamic model/neural network引用本文复制引用
基金项目
国电南瑞科技项目(524608220021)
出版年
2024