In power systems, some components such as harmonics / interharmonics and noise arise in electrical signals due to varying sources and loads, resulting in changes in the purity of the electrical signal. Therefore, continuous monitoring and accurate analysis of electrical signals become mandatory. One of the common methods in this field is the Fast Fourier Transform (FFT), which subjects electrical signals to continuous analysis through sliding windows at the intervals formed by the standard of IEC-61000-4-7. From this analysis, various parameters are derived and compared with the threshold values specified in other standard, the IEEE-1159. However, due to variational field conditions, as well as factors such as frequencies of sampling, main, and additional components, measurement window, and Signal-to-Noise Ratio (SNR), various measurement errors, occur. These difficulties further complicate the already difficult task of accurate measurement in variable conditions and can lead to errors in preventive measures and control procedures. It is therefore of great importance to gain an understanding of the effects of these parameters and to implement improvements to the methods accordingly. This study emphasizes the significance of parameter selection for FFT and presents an investigation of FFT responses in terms of different parameters. Furthermore, it presents measurement errors according to changes in the signal and presents a basic interface design showing these errors. It was found that even small changes, such as a 1/2000 change in the sampling frequency, a 0.5 Hz shift in the fundamental frequency, or a 1/1000 difference in the measurement window, resulted in significant errors. These findings highlight the necessity of judicious parameter selection for accurate computation and signal monitoring and demonstrate that improvements in the FFT method are needed to adapt to changing conditions.
Read full abstract