Improved Algorithms For Deflection Estimation Of Positioning Machine

In the control system of positioning machine, the head’s positioning accuracy is the most important factor which determines the yield of production.For such kind of high-precision devices, one of the most important factors that influence their positioning accuracy is the deflection in the mechanical structures. The finite element method(FEM) is the most common method used to investigate deflections. However, it has such a high computational complexity that it cannot be used in real-time deflection correction applications. In this paper, the on-line deflection estimation model of the X-axis under static conditions has been investigated. Firstly, the practical model of positioning machine was established in the Solid Works and the FEM simulation results were obtained through ANSYS. These enables the feasibility study and provides the source data for the following research. Secondly, a simplified static mechanical model was built, and its parameters were identified by FEM results. The proposed estimation model have a good performance on both precision and computational complexity. Thirdly, an improved estimation model was built by taking the structural strength of linear guides into consideration. It is proved that this method has a higher precision that the former one. Lastly, in order to obtain high precision results for the slope of the beam, an improved model based on local curve fitting estimation was proposed. During the realization of this novel estimation model, some difficulties such as obtaining the inverse of a problematic matrix and avoiding the Runge phenomenon were overcome. Results show that the new model is capable of providing a very accurate slope information which can be used to calculate the angle error of the positioning head.