基于动力学参数辨识的两轮机器人平衡控制

董慧芬, 姚勤燃, 王俊峰

控制工程 ›› 2025, Vol. 32 ›› Issue (2) : 345-353.

控制工程 ›› 2025, Vol. 32 ›› Issue (2) : 345-353.

基于动力学参数辨识的两轮机器人平衡控制

作者信息 +

Balance Control of Two-wheeled Robot Based on Dynamic Parameter Identification

Author information +
文章历史 +

摘要

针对非同轴两轮机器人平衡控制算法设计过程中,机器人动力学建模不精准导致控制器设计困难的问题,提出一种基于小批量梯度下降法的极大似然估计参数辨识方法。首先,用欧拉-拉格朗日方法建立两轮机器人动力学模型;然后,根据激励轨迹用极大似然估计方法进行动力学模型参数辨识;最后,用辨识参数和三维软件计算参数分别设计线性二次型调节器平衡控制算法,提升机器人倾角的控制精度。实验结果表明,该方法得到的动力学模型与真实模型的拟合度达到92.5%,基于此方法设计的控制器平衡控制效果显著,在静态和动态下均具有良好的平衡性能。

Abstract

In the design process of balance control algorithm for non-coaxial two-wheeled robot, it is difficult to design the controller due to inaccurate robot dynamics modeling. A maximum likelihood estimation parameter identification method based on small batch gradient descent method is proposed. Firstly, the Euler-Lagrange method is used to establish the dynamic model of two-wheeled robot. Then, according to the excitation trajectory, the maximum likelihood estimation method is used to identify the dynamic model parameters. Finally, the linear quadratic regulator balance control algorithm is designed with the identification parameters and the three-dimensional software calculation parameters to improve the control accuracy of robot inclination. The experimental results show that the fitting degree between the dynamic model obtained by the proposed method and the real model reaches 92.5%. The balance control effect of the designed controller is obvious, and it has good balance performance under static and dynamic conditions.

关键词

参数辨识 / 两轮机器人 / 小批量梯度下降 / 极大似然估计 / 线性二次型调节器

Key words

Parameter identification / two-wheeled robot / mini-batch gradient descent / maximum likelihood estimate / linear quadratic regulator

引用本文

导出引用
董慧芬, 姚勤燃, 王俊峰. 基于动力学参数辨识的两轮机器人平衡控制[J]. 控制工程, 2025, 32(2): 345-353
DONG Huifen, YAO Qinran, WANG Junfeng. Balance Control of Two-wheeled Robot Based on Dynamic Parameter Identification[J]. Control Engineering of China, 2025, 32(2): 345-353

3

Accesses

0

Citation

Detail

段落导航
相关文章

/