无偏差最小二乘法伺服控制系统参数辨识

为准确辨识伺服控制系统参数，针对高性能运动控制算法的要求，在分析典型伺服控制系统结构模型基础上，基于“库仑摩擦+黏性摩擦”模型，设计了无偏差最小二乘方法，研究其在“工控机+运动控制卡”数控系统平台上的实现方法。实时采集电机角速度数据，利用设计的方法估算伺服系统转动惯量、黏性摩擦和库仑转矩，通过角速度重构，对比理论计算和实际测量角速度,验证参数辨识的有效性和准确性。理论分析和实验结果表明，所设计的参数辨识方法能够准确辨识系统参数。

Parameter Identification of Servo Control System Based on Unbiased Least Square Method

Aiming at accurate parameter identification  of  a  servo  control system and the  demands of high performance motion control algorithm, a  typical servo control system model was  analyzed and the unbiased least square  method was designed based on  Coulomb friction and viscous friction model.Implementation methods of it were  studied on the hardware of industrial PC and motion control card. Motor angular  speed data was collected in real time. The inertia,viscous friction and  Coulomb torque of system were estimated using the least square method. The validity and the correctness of parameter identification were verified through reconstructing the  angular speed, comparing the actual measurement with theoretical calculation  angular  speed. Theoretical observation and experimental results  show that the parameters of the servo system can be accurately identified in this kind of CNC system platform with the unbiased least square method.