摘要
根据Lu Gre摩擦模型辨识理论,以伺服电机驱动轴与伺服电机内部的摩擦模拟机器人关节摩擦,建立Lu Gre摩擦模型。对伺服电机做基于固高卡的模拟量控制,通过C++编程,从编码器中读取角度、转速值、加速度值。正转速度和对应的驱动力矩,以及反转速度和对应的驱动力矩,分别构成静态参数辨识的两组数据。位移、速度、加速度和驱动力矩构成动态参数辨识数据。在MATLAB中编写粒子群算法辨识程序,对以上数据进行处理,最终得到Lu Gre模型的6个参数的辨识值。
The friction between the driving shaft of a servo motor and its stationary part is simulated as friction of a robot joint and its Lu Gre friction model is built based on the identification theory of the Lu Gre friction model. The modulating control of the servo motor through Googol card is implemented using the visual C＋＋ program, the values of angles, rotational speed of the motor and acceleration values can be obtained from the encoder. The values of positive-rotating speed and those of the corresponding drive torque, as well as the values of the reverse-rotating speed and those of the corresponding drive torque are used as the two sets of data to identify the static parameters. The values of angle, velocity, acceleration and driving torque are used as the data to identify the dynamic parameters. A PSO identification program to process the data mentioned above is made in the environment of MATLAB, and the six parameters of the Lu Gre friction model are obtained.
出处
《微型机与应用》
2014年第19期76-78,82,共4页
Microcomputer & Its Applications
关键词
LU
GRE
摩擦模型
模拟量控制
粒子群算法
Lu Gre
friction model
modulating control
particle swarm optimization
