電子產品中絕大部分都有IC (integrated circuit)的存在，隨著IC晶片逐漸朝向輕量化、薄型化與高密度發展，為了維持高良率與應有的生產效益，晶圓片切割技術也隨著基本材料的不同以及晶圓片的薄化，在封裝製程上已被視為控制品質成功關鍵的因素之一。然而，以傳統試誤法與工程師自身經驗來決定的切割參數，並無法保證這些切割參數值為最佳設定。本研究結合實驗設計、基因規劃及人工免疫演算法提出一個系統化的晶圓片切割品質最佳化的參數設計程序。同時，透過一個改善薄型晶圓片－蕭特基二極體(Schottky diode)切割製程的實證案例，以驗證所提之演算法的可行性和有效性。驗證結果顯示出薄型晶圓片的切割品質均能接近理想值並完全符合其規格限制。因此，本研究所提的整合式演算程序在解決單一品質特性參數設計問題上是一種有效的工具。

All electronic products almost contain integrated circuits (IC). As the progress of the lightening, thinning and high density of IC, the technologies of wafer dicing has been one of the critical factors which influence the slicing capability. Traditionally, engineers determined the optimal settings of slicing parameters through trial and error or according to their own experience. However, the parameter setting cannot be proven to be real optimal. This study integrated the experimental design, genetic programming and artificial immune algorithms to develop a systematic procedure for optimizing parameter settings in wafer slicing. A case study of improving the slicing process of Schottky diodes was used to demonstrate the feasibility and effectiveness. The experimental results revealed that the cracks in both front and back sides of wafers can fulfill the specifications. Hence, the proposed procedure can be considered an effective method for resolving a parameter design problem with single quality characteristic.