A COMPARATIVE STUDY AND PERFORMANCE ANALYSIS OF COMMONLY USED SIGNAL PROCESSING TECHNIQUES IN POWER QUALITY APPLICATIONS

Abstract

It is quite important to identify the power quality problems that may occur in the process from generation to distribution of electricity. The voltage obtained from an ideal three-phase alternating current source is a continuous sinusoidal voltage with approximately the same amplitude and 120° phase difference between them. Current or voltage distortions that may occur in power systems such as harmonics and the voltage sag, swell, and interruption cause the low quality of the power. These power quality disturbances may cause the failure of the devices used by the consumer, an increase in maintenance costs and current or voltage unbalance. The causes and types of problems must be well known to improve power quality. Especially if the type of power quality disturbances is classified correctly, the effects of disturbances under load can be identified, the source of the disturbances can be analyzed and thus appropriate solution methods can be developed. It is obtained that the increasing system reliability, reducing line losses, and increasing energy efficiency thanks to improving the power quality. Power quality problems are generally described as waveform distortions in a current and voltage. Sinusoidal distortions in voltage and current waveforms in the electric grid create harmonics. In this study, the purification of harmonics is provided by using signal processing techniques commonly used in different applications in the literature. The optimum signal processing technique to be used in solving power quality problems is determined It is presented comparatively simulation results of signal processing techniques that Fourier Transform (FT), Short-Time Fourier Transform (STFT) and Wavelet Transform (WT). the.  The analysis results show that WT is superior to other used transform techniques for signal processing. The signals can be analyzed locally by obtaining low-frequency information in long-time intervals and high-frequency information in a short time interval thanks to the WT.