A STABILIZED GENETIC ALGORITHM FOR EVALUATING THE ECONOMIC VIABILITY OF DYNAMIC DISTRIBUTION NETWORK RECONFIGURATION

By Pham Hong Thanh
DOI: 10.37550/tdmu.EJS/2025.04.689

High penetration of photovoltaic (PV) sources causes volatility in distribution networks, challenging conventional operational strategies. This study introduces a multi-objective optimization framework using a Stabilized Genetic Algorithm (SGA) that co-optimizes daily energy losses and switching asset depreciation over typical and extreme loading scenarios. Contradicting common assumptions, results show that zero switching operations, i.e., maintaining a robust static configuration - yield optimal economic outcomes for the IEEE 33-bus test system, regardless of switching cost magnitude. The work formalizes an economic viability threshold for DDNR, providing network operators with a quantitative tool to assess when dynamic reconfiguration is truly justified. Results reveal that for the IEEE 33-bus system with PV integration, a robust static configuration remains economically optimal regardless of switching cost magnitude. The primary contribution is the formalization of an "Economic Viability Threshold" framework, providing DNOs a quantitative tool to determine when DDNR is truly justified. This framework provides a crucial, data-driven tool for network operators to prevent unnecessary investment in complex control schemes, ensuring that grid modernization efforts are both technically sound and economically viable