A hybrid approach based on IGDT‐MOCMA‐ES method for optimal operation of smart distribution network under severe uncertainties

Abstract

Uncertainties involved in renewable generation and electrical demand pose significant technical challenges with the concomitant financial consequences in smart distribution networks (SDNs), particularly in the current electricity market, which is restructured and features smart grids. The present paper introduces a decision-making tool based on a risk-averse strategy to help with the smart distribution network operator (SDNO) in day-ahead operational practices, including optimal unit commitment (UC) and optimal distribution feeder reconfiguration (DFR). The tool is meant to reduce electricity prices presented to the electrical consumers and to optimize financial transactions with the energy market, distributed generation (DG) reliability, electricity storage system (ESS) dispatch, and planning interruptible electrical demands to secure specified revenue targets for SDNO with the risk-averse strategy. A bi-level stochastic optimization problem based on Information gap decision theory (IGDT) is considered to keep the SDNO from risks inherent in the information gap present between the predicted and actual uncertainty variables. The bilevel stochastic optimization problem is reorganized into a single-level problem obtained by Karush-Kuhn-Tucker method. As uncertainty variables compete to expand their enveloped-bounds, multi-objective covariance matrix adaptationevolution strategy (MOCMA-ES) is employed to address the multifaceted IGDTbased stochastic optimization problem proposed in the study. Finally, the efficiency and efficacy of the suggested model are appraised on an IEEE 33-bus SDN. Simulation results show that optimal UC with both DFR and demand response program increases the total revenue by 8.1% compared to optimal operation without them.

Multi-Objective Optimal Allocation of Hybrid Photovoltaic Distributed Generators and Distribution Static Var Compensators in Radial Distribution Systems Using Various Optimization Algorithms

“…where, σ and μ the standard deviation and mean, respectively [34]. The solar irradiance state's probability s through any certain hour could be formulated as:…”