本文提出田口方法應用於被動式諧波濾波器最佳化設計。直交表已廣泛應用於解決實際問題的之重要實驗程序，主要原因是直交表具有幾何平衡特性，使用較少的實驗獲得更可靠的濾波器最佳化組合。配合信號雜訊比(S/N)，衡量最佳化濾波器品質的穩健性；濾波器設計時，有些不確定因素，例如系統阻抗變化及不同負載產生諧波電流，以雜訊因子來考量，以探討其對濾波器最佳化設計的影響程度，並藉此瞭解何種濾波器組合，可以對諧波有最佳濾波效果。為驗證所提方法有效性，本文以某化工廠為實例，模擬結果顯示將田口方法應用於被動式諧波濾波器設計能有效降低系統之諧波失真，提昇供電系統的電力品質。

This report presents the Taguchi method, used to investigate the planning of large-scale passive harmonic filters. Orthogonal arrays are widely employed in experimental processes for solving practical problems encountered because such arrays are geometrically balanced in their coverage of the experimental region after only a few representative experiments have been implemented. The signal-to-noise ratio(S/N) is employed to investigate the robust design of harmonic filters. In the design of harmonic filters, some uncertainties should be considered. Noise factors have been considered to cause filter parameter detuning, loading uncertainty and changing of system impedances. Our design approach has been applied to harmonic mitigation problems for a chemical plant, where three large-size 6-pulse rectifiers are installed. The purposes are to minimize the total demand distortion of harmonic currents and total harmonic distortion of voltages. Fil-ters loss, reactive power compensation, and constraints of individual harmonics should are also considered. Additional results related to the Taguchi method are also reported and discussed as well.