晶圆校准器误差补偿与拟合算法的研究OA北大核心CSTPCD

Wafer alignment calibrator plays a crucial role in semiconductor manufacturing and inspection processes.However,the accuracy of wafer center positioning can be affected by measurement errors.Consequently,the development of error compensation algorithms is paramount to refining the accuracy of wafer alignment calibrator.To address the impact of wafer carrier surface tilt on sampling accuracy,a physical model of a tilted wafer and its elliptical projection on the horizontal plane is established.An analytical relationship between the ellipse and the wafer center is derived from this model.Subsequently,an elliptical equation is employed for fitting purposes.To further enhance compensation and fitting accuracy,an extended Kalman filter-based ellipse fitting algorithm with a forgetting factor is proposed.This algorithm employs density-based spatial clustering of applications with noise for data preprocessing,coupled with improved progressive consistent sampling to improve the distribution of sample points.Additionally,an iterative reweighted least squares method is applied for ellipse fitting,resulting in high-precision estimation of the wafer center.Experimental results demonstrate that the proposed method achieves a fitting accuracy of 99.95％after error compensation,thus validating its effectiveness.