多自由度柔性关节机器人奇异摄动控制若干问题研究

项目名称： 多自由度柔性关节机器人奇异摄动控制若干问题研究

项目编号： No.61203337

项目类型： 青年科学基金项目

立项/批准年度： 2013

项目学科： 自动化学科

项目作者： 刘华山

作者单位： 东华大学

项目金额： 25万元

中文摘要： 针对传统的柔性关节机器人简化模型不能完全反映其柔性传动特性、关节电机或连杆的某些状态量无法测量或易混入噪声、关节电机存在力矩饱和现象等问题，本项目以奇异摄动理论为基础，借助滤波估测和有界控制等手段，并加以实验验证展开应对研究。首先，考虑柔性关节非线性刚度扭矩、柔性扭曲、摩擦损耗和惯性交叉耦合等因素，对柔性关节机器人系统的刚性和柔性部分分别建模，以连杆角度变量作为慢变量，关节弹性力作为快变量，完成模型的奇异摄动转换；然后，采用扩展卡尔曼滤波（EKF）与无损卡尔曼滤波（UKF）实现仅需机器人连杆和电机角度位置测量的输出反馈控制及仅需电机角度位置和角速度测量的部分状态反馈控制；最后，基于一类具有平滑趋近饱和特性的函数，提出一种可适用于一般多自由度柔性关节机器人的有界控制律。本项目将进一步完善柔性关节机器人的奇异摄动控制研究，并为其工程应用提提供理论依据和方法指导。

中文关键词： 机器人；柔性关节；奇异摄动；卡曼滤波；跟踪控制

英文摘要： For the robot manipulators with flexible joints, traditional simplified model can not fully reflect the flexible transmission characteristics, some state variables of joint motors or links can not be measured or usually mixed with noises, the torque outputs of the joint motors are saturated. To deal with these issues, theoretical study based on the singular perturbation theory, filter estimation and bounded control, are implemented with experimental verification. Firstly, considering the nonlinear stiffness torque, soft windup, friction losses, and inertial cross-coupling factors of the flexible joint , the models of the rigid and flexible parts of the robot are built respectively. Then taking the link angle as the slow variable, and the joint elastic force as the fast, singular perturbation conversion of the whole model is fulfilled; Secondly, extended Kalman filter (EKF) and Unscented Kalman filter (UKF) are applied to achive the output feedback control with only angular position measurements for links and motors, as well as the partial state feedback control with only angular position and velocity measurements for motors; Finally, a class of functions, characterized by smooth approaching to the saturation, are designed to propose a bounded control law for general multi-degree of freedom robot manipulators

英文关键词： robot；flexible joint；singular perturbation；Kalman filter；tracking control