本研究使用L9田口法實驗，以紅銅做為放電電極對冷作工具鋼SKD11進行放電加工實驗，探討放電參數(低壓電流、引弧電流、加工深度、放電幅、休止幅)對電極邊角磨耗、電極消耗率、工件邊角半徑、工件材料移除率和工件表面粗糙度的影響。實驗得知，當低壓電流大於4 A後，電極消耗率及材料移除率會隨著加工深度增加而些微下降。電極邊角磨耗和工件邊角半徑會隨著加工深度增加而增加，但其增加量會隨加工深度而趨緩。當放電幅低於30 μs時，放電幅的影響較低壓電流明顯，工件表面粗糙度會隨著放電幅的提升而增加，但當放電幅高於30 μs，低壓電流的影響會大於放電幅，工件表面粗糙度會隨著低壓電流的提升而增加。根據訊號雜訊比的分析發現在不同加工深度下，低壓電流及放電幅為影響工件表面粗糙度的主要參數。隨著加工深度增加，低壓電流及放電幅對電極邊角磨耗及工件邊角半徑的影響程度會增加。在不同加工深度下，低壓電流為主要影響電極消耗率和工件材料移除率的加工參數。最後本研究利用倒傳遞神經網路對實驗結果進行模擬預測，將模擬預測值與實驗結果作比較發現，雖然各結果對於隱藏層及神經元數量需求不同，但只要能提供足夠的資料使神經網路做學習，便能讓神經網路擁有良好的預測能力，使預測結果和實驗相比能在9 %的誤差值內。

In this study, the L9 Taguchi method was used to perform the electric discharge processing on SKD11 with copper as electrode tools. The effects of different processing parameters (pulse current, arc current, open voltage, pulse on-time and pulse off-time) and processing depth were observed with material removal rate, electrode tool wear rate, electrode tool’s corner radius, material’s corner radius, and surface roughness. The result showed that when pulse current exceeds 4 A, electrode tool’s wear rate and material removal rate decrease slightly as processing depth increases. Electrode tool’s corner radius and material’s corner radius increase as processing depth increases, but the amount of radius change gradually decreases with processing depth. When pulse on-time less than 30μs, surface roughness increases as pulse on-time increases, but pulse on-time higher than 20μs, surface roughness increases as pulse current increases. In view of the electric-discharge processing results observed in the experiment, due to the use of Taguchi permutations and combinations, it is necessary to calculate the signal to noise ratio, that is, the S/N ratio, to discuss the machining results of the performance, finally calculate the percentage of contribution to find the highest impact factor of the processing parameters. In the result of the electric discharge processing, pulse current and pulse-on-time are the maximum influence factor in surface roughness. Pulse current is the maximum influence factor in electrode tool’s corner radius and material’s corner radius. The percentage of influence of pulse current and pulse on-time to electrode tool’s corner radius and material’s corner radius increase with processing depth.