S-Transform Based Novel Indices for Power Quality Disturbances

Abstract

Power quality (PQ) has recently become an increasing concern for electric utilities and their customers due to the ever-growing proliferation of power electronic devices and nonlinear loads in electrical power networks. The opening of power markets, and the deregulation and restructuring throughout the world are further changing the framework in which power quality is addressed (Beaulieu et al., 2002). Therefore, the techniques for power quality study and power disturbance mitigation are capturing increasing attention, and consequently, manufacturers are integrating power quality monitoring functions in their products such as power meters, digital relays and event recorders (Ward, 2001). One of the most important aspects in power quality analysis is to evaluate the extent of power disturbances and their negative impacts on power systems. Power quality indices, as a powerful tool for quantifying power quality disturbances, are the concise numerical representations characterizing the nature based on the time and frequency information of the disturbance waveform (Lin & Domijan, 2005). They also serve as the basis for illustrating the negative impacts of electrical disturbances and for assessing compliance with the required standards or recommendations within a given regulatory framework. A power quality disturbance usually involves a variation in the electric voltage or current and can be classified based on the waveform time-statistical characteristics into two groups: steady disturbances (harmonic distortion, unbalance, flicker etc.) and transient disturbances (voltage sags, voltage swells, impulses, oscillatory transient etc.). Several power quality indices identified and utilized in the past were accepted and worked well for single and three-phase balanced systems and with periodic stationary waveforms, however, pitfalls of these indices in common use are discussed in (Heydt, 1998, 2000). With the severe changes in the waveforms of the current and voltage signals in the power network, the traditional definitions of the previous indices are no longer valid. As transient disturbances are characterized by spectral components that are significantly time varying in amplitude and/or in frequency; it is important to accommodate the time information in the indices calculation so that they can provide more sensitive interpretation of the disturbances. Recently, one of the most interesting applications is the assessment of power quality by redefining new indices for transient disturbances using signal processing techniques. New power quality indices have been suggested based on the short time Fourier transform (STFT) (Jaramillo et al., 2000), time frequency distribution(TFD)(Shin et al., 2006) and stationary wavelet transform(SWT)(Morsi & Hawary, 2008). The integrating of the time