Signal processing and AI based diagnosis of power quality disturbances: A review

The precise diagnosis of power quality disturbances (PQDs) has now become a significant concern among the utility engineers as well as consumers due to the high cost of downtimes associated with it. Numerous methods, such as, Artificial Intelligence (AI), Signal Processing (SP), Space Vector Representation, Symmetrical Component, have been evaluated for the precise diagnosis of PQDs. Among these methods, AI and SP based techniques have received extensive attention by the researchers and industry engineers. This paper discusses the various AI and SP based methodologies currently used for the diagnosis of PQDs. Existing AI and SP based methods are critically reviewed to highlight their applications, merits, and shortfalls. It is revealed that, besides the numerous applications and merits of these methodologies, none of them is found proficient for the precise diagnosis of PQDs. Accurate recognition of PQDs is still a challenging task. Therefore, the need of incorporation of new techniques for the accurate estimation of PQDs has been asserted.