在電力系統中，諧波與閃爍是極為普遍的電力品質擾動現象。為了改善它們對電力系統所形成的汙染問題，必須有一套快速與準確的量測方法。而目前離散傅立葉轉換(DFT)為基礎的演算法，仍是各種量測儀器最常用的頻譜分析工具。但是系統頻率變動所形成的洩漏效應問題，一直是DFT在準確度上所面臨的嚴重挑戰。近年來，連續小波轉換(CWT)為基礎的演算法雖可改善此洩漏的問題，但由於計算時間過久，使其在即時量測上之可行性受到質疑。故本文根據理論上DFT的演算過程，分析一般系統頻率變動對DFT用於諧波及閃爍量測計算上所形成的振幅及相角誤差，以瞭解在正常系統頻率變動範圍內影響的程度，作為選擇適當DFT演算法之明確參考。此外在仔細比較多種頻率估測演算法後，採用Prony為基礎的演算法來量測系統頻率，納入CWT演算法之中，以提升閃爍演算法速度。同時，由於理論分析結果顯示，系統頻率變動對諧波量測影響不大，故仍以FFT作為諧波算式，因而形成FFT-Prony-CWT混合式諧波與閃爍量測演算法。經模擬及實測驗證，此混合式演算法除擁有比全連續小波轉換演算法具有較快的計算速度外，也不失量測的準確度。 Harmonic and flicker are common causes resulting in power quality related problems, their instrumentation and measurement schemes play a critical role on the solutions. The discrete Fourier Transform (DFT) based algorithms have been very popularly used to calculate the amplitude and the phase spectra of a harmonic distorted and flicker-modulated signal. But, the leakage effect due to system frequency variation is still a troublesome problem for accurate measurement. Therefore, continuous wavelet transform (CWT) based algorithm was proposed to mitigate the leakage effect; however, the feasibility of the algorithm is doubtable for real-time measurement because of their tremendous calculation time. Accordingly, this thesis theoretically derives the leakage effect from the development process of DFT for harmonic and flicker measurements respectively. So, we can have a definite concept about the degree of the leakage effect under the possible range of system frequency variation as an appropriate DFT-based algorithm is determined. Besides, the Prony-Based (P.B.) algorithm is incorporated into the CWT-based flicker measurement algorithm for the system frequency estimation after many related algorithms are compared in their calculation time and accuracy under different waveform distortion conditions. Since the leakage effect of the system frequency variation on the DFT-based harmonic measurement can be neglected, the Fast Fourier Transform (FFT) is adopted to become a hybrid algorithm (FFT-Prony-CWT) for the simultaneous measurement of harmonic and flicker. The simulation and field signal test results show that the calculation of FFT-Prony-CWT hybrid algorithm is faster than full CWT-based algorithm while the accuracy is still reserved.