由於科技不斷地進步,從傳統工業轉型成高科技產業,導致用電量隨之上升,所以供電品質與電力品質之可靠度極為重要,相對地,隨之衍生而來之電力品質問題便層出不窮。而引起電力品質之不良因素頗多,其中較為常見之因素為:電力諧波、電壓驟降、電壓閃爍、電壓陡昇、電壓中斷、三相不平衡、電壓突波與電流突波等,這些不良因素可能影響廠內設備之正常運轉,且對電機與電子設備造成破壞,甚至危害操作人員之安全。 本論文除使用短時傅立葉轉換(Short Time Fourier Transform,STFT)及一般性時-頻傅立葉轉換(Generalized Time-Bandwidth Product (GTBP) Optimal STFT)外,並針對上述兩方法之缺點,提出一修正型傅立葉轉換(Modified STFT)方法。而對數種不良電力品質訊號之模擬與探討,可驗證本論文所提之方法較其它方法具有更佳之準確性。
Because of the progress constantly of science and technology, it makes the transition from traditional industry into a high-tech industry. Therefore, this causes the rising of the power consumption, and the reliability of supply power quality is extremely important. Besides this, there are more and more problems about power quality. The bad factors which influence the power quality are quite many. They include harmonics, voltage sag, flickers, voltage swell, voltage interruption, three-phase unbalance, inrush voltage and inrush current. These bad factors may affect the normal operation of the equipment in the factory, and cause the damage of the electrical equipments, even endanger the operator security. In this thesis, besides using the Short Time Fourier Transform (STFT) and the Generalized Time-Bandwidth Product (GTBP) optimal STFT, the Modified STFT method to improve the shortcoming of the two above methods is proposed. The simulation and discussion of many kinds of power quality signals can demonstrate that the proposed method has better accuracy than other methods.