故障電流中的衰減直流成分會造成測距電驛在測量上有兩大影響,其一是測距電驛阻抗會振盪,無法立刻收斂至正確值。其二是發生比流器飽和,會有電流波形失真的現象發生。針對以上的兩個問題,本論文提出之演算法可直接濾除故障阻抗中的衰減直流成分以得到穩態測距結果。另一方面,本論文利用計算得到的阻抗變動來偵測比流器飽和之發生,並使用短視窗離散傅立葉轉換演算法得以加快偵測速度。本論文之理論驗證分成軟體模擬及硬體實作兩大部分,軟體部分,我們使用SIMULINK模擬出故障時的電力訊號並作為演算法運算初步的資料來源。硬體部分,我們把SIMULINK模擬的資料匯入數位訊號處理器(DSP,Digital Signal Processors),搭配Code Composer 3.1(CCS 3.1)之整合發展環境軟體,同樣也可正確的估算測距阻抗以及偵測比流器飽和。由實驗結果顯示,MATLAB模擬和DSP實測結果極為相同,這表示提出之演算法可應用於DSP實作中。
The decaying DC components in fault currents will put two major impacts on distance relays. The first impact is that the distance relay cannot obtain accurate results immediately due to their impedance oscillations. The second impact is that the saturated Current Transformer (CT) will cause the distorted current measurements. In considering the above two problems, this thesis presented a new algorithm which can directly remove the decaying DC components from fault impedances. On the other hand, the variations of the obtained impedance can be used to detect the CT saturation. Meanwhile, the Short Window Discrete Fourier Transform (SWDFT) was adopted to speed up the rate of detection. In this thesis, software and hardware environments were used for the theoretical evaluations. In the software environment, some SIMULINK-based model were used to generate the fault signals for computations, In the hardware environment, the fault signals generated from SIMULINK were sent into a Digital Signal Processor (DSP) cooperating with Code Composer 3.1 (CCS 3.1) to obtain the fault impedance and detect the CT saturations. The results show that both the software and hardware environment are able to achieve similar results. It is shown that the proposed algorithms have the potentials to be established on a practical DSP environment.