AI-Based Multi-Objective Distribution Network Reconfiguration Considering Optimal Allocation of Distributed Energy Storages and Renewable Resources

Abstract

This paper presents a innovative approach integrating Information Gap Decision Theory (IGDT) with multi-objective optimization for distributed energy resource placement and network reconfiguration. The research introduces a dual-mode optimization framework addressing both grid-connected and islanded operations, extending beyond traditional single-mode analyses. The methodology employs a three-tier approach: optimal DER placement for loss minimization, strategic Energy Storage System deployment for operational resilience, and dynamic network reconfiguration. The primary technical contribution is an advanced AI-based optimization algorithm that synthesizes backward-forward load flow analysis with market dynamics, achieving 27% improved computational efficiency. The algorithm incorporates stochastic variables for renewable generation uncertainty through IGDT framework, ensuring system stability under varying intermittent resource penetration. A key innovation is the multi-objective function optimizing technical and economic parameters, including power loss reduction, voltage profile enhancement, and carbon emission minimization. The research introduces dynamic network reconfiguration responding to both technical limitations and market signals, demonstrating 15% improved loss reduction compared to static configurations. Validated on a modified IEEE 33-bus system, the methodology achieved 32% reduction in power losses during grid-connected operation and 40% decrease in ENS demand during islanded operation, while maintaining voltage profiles within ±5% of nominal values. This research establishes a new paradigm bridging theoretical optimization and practical implementation constraints.