Real-time Estimation of High-precision Actuator’s Variable Stiffness for Vibration Isolation

Certain flexure-guided high-precision actuators have a variable stiffness. By tuning it, quasi-zero stiffness can be realized to improve vibration isolation for high-precision motion control. To enable such high-precision actuators to perform optimally in various environments, this paper presents the development of an estimator for a variable stiffness. As a laboratory setup for experimental validation, a flexure-guided hybrid reluctance actuator (HRA) is developed with feedback control for high-precision motion control with a bandwidth of 721 Hz. The HRA can vary its variable stiffness by adjusting its mover position. To estimate the variable stiffness, an extended Kalman filter (EKF) and an unscented Kalman filter (UKF) are designed individually for comparison. To experimentally evaluate the stiffness estimation performance with the EKF and the UKF individually, the variable stiffness of the HRA is identified by frequency response analysis and varied between -7.68 kN/m and 34.5 kN/m. Experimental results successfully demonstrate that both the EKF and the UKF estimate the variable stiffness with an averaged estimation error less than 1.1 kN/m, revealing that they have almost the same estimation performance and the real-time estimation capability when the stiffness gradually changes.