Dynamic Position Detection Method for Large-Diameter Rotary Shaft Based on MIFA-AFKF

Abstract

The dynamic position variables of large-diameter rotary shaft are not easy to measure precisely or cheaply, however, the accuracy of dynamic position is of great significance to the movement controlling of key rotary shaft. In this paper, a novel real-time detection method is presented, which can be used to estimate the dynamic position of the large-diameter rotary shaft rotating with constant velocity, constant acceleration or variable acceleration. In order to estimate the position variables, multi-sensor information fusion algorithm based on adaptive fading Kalman filter (MIFA-AFKF), which combines conventional information fusion algorithm with an adaptive fading factor, is proposed to fuse multi grating-sensor information. The method based on MIFA-AFKF is demonstrated by Monte Carlo simulation, and also used to measure the angle position of the hollow shaft at the machine head and machine tail of a horizontal riveting machine. The two results show that the algorithm can significantly improve the dynamic position accuracy.