本文聚焦於大口徑非球面拋光製程，針對CNC數控拋光機台建立迭代式進給速率演算法，此演算法基於進給速率與材料移除深度之關聯性建立模型，基於迭代法計算出用以收斂目標面形誤差所需之進給速率分佈，並可進一步預測使用此進給速率分佈進行拋光加工後之表面殘差；計算結果顯示於∮100 mm之範圍內可將PV值收斂至1 nm以內，於∮90 mm的範圍內則可收斂至0.5 nm以內，證明應用本研究所建立的迭代式進給速率演算法可於拋光製程中提供相當高的拋光預測度。

This study established an iterative algorithm, which based on feed rate, for the CNC bonnet polishing process. This algorithm calculates the corresponding feed rate distribution which eliminates the surface form error based on the relationship between the feed rate and the material removal depth, furthermore, the predicted surface residual will also be calculated and demonstrated. The calculation result shows that in the range of ∮100 mm, the PV value of surface residual could be converged under 1 nm, and the PV value within ∮90 mm range could be further converged under 0.5 nm, which sufficiently supports the high convergence of the algorithm constructed in this research.