Includes a title page in Arabic

Thesis (M.Sc.)
Master Egypt - Japan University of Science and Technology (E-JUST) - School of Electronics , Communication , and Computer Engineering - Electrical Power Engineering Department 2024

Phase identification involves determining which phase each end-customer is connected to in a multi-phase distribution network Knowing the correct customer-phase connectivity provides Distribution System Operators (DSOs) with critical information for efficient operation and management of Low-Voltage Distribution Networks (LVDNs) Additionally , accurate phase connectivity information is essential for increasing the hosting capacity of Distributed Energy Resources (DERS) , which are key players in modern power systems On the other hand , incorrect customer-phase connectivity increases the risk of phase unbalance , leading to nuisance tripping of protection devices and equipment damage for assets such as transformers Usually , the phase connectivity information of a LVDN is known to DSOs at the time of planning and commissioning the power line However , due to the complexity and the various changes that occur during its operation , such as new customer connections , maintenance and repair operations , cyber attacks , emergency restoration services , etc. , the original data files maintained at the power management center often contain false connectivity information This challenge is traditionally addressed by applying hardware-based phase identification techniques Nevertheless, such approaches are costly , prone to human errors , and time - consuming, as they involve either installing expensive high-precision devices or employing field-based methods To overcome the above challenges , this thesis develops a novel data- driven method to identify the phase connectivity of end-customers using Advanced Metering Infrastructure (AMI) voltage and current measurements , collected every 15 minutes Firstly , a preprocessing method that employs linear interpolation and singular value decomposition is adopted to improve the quality of the smart meter data Secondly , using Kirchoff's current law and correlation analysis , a 0-1 linear integer programming optimization model is built to uniquely identify the phase to which each customer is connected The datasets utilized are obtained by performing power flow simulations on a modified IEEE-906 test system using OpenDSS software The robustness of the proposed identification algorithm is tested against dataset size , missing data , measurement errors , and the influence of rooftop photovoltaic generation systems To explore the benefits associated with knowing the correct phase identification information

Issued also as a digital file (for more information please check our Digital Repository)

AnD ACM