加工產品的製程中，加工參數對於產品成品的加工品質有相當大的影響，通常需要選擇一組適合的參數來達到指定的加工品質。這樣的工作往往需要仰賴少數有經驗的老師傅或一份鉅細靡遺的加工操作手冊才能進行，要不然就只能透過反覆的試誤來試出希望的組合。隨著人工智慧時代的來臨，本文介紹透過強化學習 (reinforcement learning) 實現了自動加工參數調整的方法來找到需求的加工品質。此外，本文也採用多目標演算法提供了一個針對多性能指標的性能偏好選擇方式，並將結果實際在深孔放電加工 (hole drilling electric discharge machining, HEDM) 機台上驗證。

During the process to machining a product, machining parameters have huge impacts on the quality of the product. We need to adjust to a proper parameter set to obtain a desired quality. This parameter tuning task often relying on a few experienced engineers or an elaborate operating manual. Otherwise, it is only possible to find the desired parameter set through try-and-error experiments. With the advent of the artificial intelligent (AI) era, we have implement an automatic parameter tuning method with reinforcement learning to obtain a desired machining quality. The method also offers a way to choose the quality preferences under multiple performance indexes, and the result is validated on a hole drilling electric discharge machining machine.