Document Type : Research article
Highlights
• Proposing a novel method for validating the radiality of distribution networks.
• Employing deep reinforcement learning as a model-free approach to solve the
distribution network reconfiguration problem.
• Investigating the impact of sensitivity analysis in hyperparameters of deep
reinforcement learning agents.
• Validation of the proposed method on well-known IEEE distribution
networks, such as 33- and 69-bus.
• Validating the proposed algorithm under uncertainties arising from varying
load conditions and distributed generation units.