近來由於石化能源及京都議定書之約束,除積極尋找替代能源外,可減輕二氧化碳排放壓力的再生能源之推廣應用,日益顯的重要,太陽能發電與風力發電等小型分散式系統與市電併聯運轉的數量乃逐年成長。 與市電併聯之分散式發電系統常出現孤島運轉現象,所謂孤島運轉現象係指分散式發電系統與市電系統並聯供電時,當市電系統發生故障或因維修與分散式發電系統解聯,但分散式發電系統未即時檢知並切離市電系統,而呈現獨立供電的現象。當發生孤島運轉現象時,孤島運轉監測系統應立即監測出此狀況並且立即停止供應電力予用戶端負載,以免造成危害。本文建立一風力發電系統,針對主動式頻率飄移,滑差式頻率位移及電壓差分相關性擾動法等三種偵測技術,針對其無法偵測區域、偵測時間等效能進行評估。 為了驗證上述三種偵測方法之效能,本文採用電壓源電流控制之反流器,並以數位訊號處理器作為系統整體控制核心,使反流器具有偵測及反孤島運轉能力,本文透過實驗驗證反孤島偵測效果並記錄其變動特徵,若以偵測時間論其效能,結果發現主動式頻率飄移與滑差式頻率位移方法較為快速,卻存在無法偵測區域,反之,電壓差分相關性擾動法之偵測方式時間較慢,卻較為可靠。
Over the last years, due to limitation of the fossil fuel resources and Kyoto Protocol of Ratification, promotion of the renewable energy applications appears to be more and more important to reduce the CO2 emission, besides developing substitute energy resources. Therefore, interconnected with utility grid systems, the amounts of dispersed generation systems, such as photovoltaic and wind generations, increase yearly. The islanding operations usually occur with the grid-interconnected dispersed generation systems. The islanding operation is so-called a situation where the dispersed power systems keep supplying electric power to a distribution network even when the utility grid is de-energized due to fault or maintenance. When the islanding operation occurs, to avoid any hazard thus caused, the islanding detection systems should detect it and disconnect dispersed generation systems from the grid immediately. In an attempt to evaluate the existing active detection methods of Active Frequency Drift method, Slip-mode Frequency-shift method and Differential Voltage Correlation methods, this thesis establishes a dispersed wind power generation systems. Evaluated is the performance of the three methods on non-detection zone and detecting time needed. In order to verify the performance of the three active methods, as mentioned above, Current-control voltage source inverter and digital signal processor (DSP) are adopted. Using the DSP as a control core, the inverter thus has the capabilities to detect the islanding operating. Through testing and recording the anti-islanding characteristics from the experiments, the Active Frequency Drift method, Slip-mode Frequency-shift method are found to be superior in terms of detection time. However, non-detection zones exist for these two methods, whereas the Differential Voltage Correlation method seems more reliable but with longer detection needed.