Bi-Level Phase Load Balancing Methodology with Clustering-Based Consumers’ Selection Criterion for Switching Device Placement in Low Voltage Distribution Networks

In the last years, distribution network operators (DNOs) assumed transition strategies of electric networks (EDNs) towards active areas microgrids where, regardless operating regimes, flexibility, economic efficiency, low power losses, and high quality are ensured. Artificial intelligence techniques, combined with smart devices real-time remote communication solutions enormous data amounts, can represent starting point in establishing decision-making to solve one most important challenges related phase load balancing (PLB). this context, purpose paper is prove that a strategy based on limited number PLB installed at consumers (small implementation degree) leads similar technical benefits as case full EDNs. Thus, an original bi-level methodology, considering clustering-based selection criterion for placement switching devices, was proposed. A real EDN from rural area belonging Romanian DNO has been considered testing proposed methodology. An degree by 17.5% represented optimal solution, leading faster computational time 43% reducing operations 92%, compared (100%). The performance indicators unbalance factor energy-saving highlighted efficiency