UKF based filtering and curve fitting for estimating the joint stiffness of stewart platform

摘要

In order to improve the precision of most control method for the stewart platform, this paper gives a novel method to determine the joint stiffness to avoid the negative effect of the actuated joint deformation. We establish the transfer function of the linearized model between the theoretical and the real value of joint deforming length so as to estimate the joint stiffness. At the same time, in view of the special noise of the distance measuring sensor output, the unscented Kalman filter is used in signal processing. The experimental results show that this method can obtain the joint stiffness parameters effectively.