A single and multiobjective robust optimization of a microgrid in distribution network considering uncertainty risk

Area of Science:

• Electrical Engineering
• Renewable Energy Systems
• Optimization Theory

Background:

• Microgrids (MGs) integrating photovoltaic (PV) and wind turbine (WT) sources with battery storage are crucial for modern power systems.
• Uncertainty in renewable energy production and network demand poses significant challenges to MG operation and economic viability.
• Robust optimization techniques are needed to ensure reliable and cost-effective MG performance under uncertain conditions.

Purpose of the Study:

• To perform single and multi-objective robust optimization of a microgrid (MG) with PV, WT, and battery storage.
• To minimize energy losses, electricity purchase costs, and power purchase costs from the MG considering uncertainty risk.
• To evaluate the system's robustness against forecasting errors in renewable production and network demand.

Main Methods:

• Implemented deterministic and robust optimization approaches using a flow-direct algorithm (FDA).
• Utilized Information Gap Decision Theory (IGDT) with a risk-averse strategy for robust optimization.
• Analyzed single and multi-objective cases to determine the maximum radius of uncertainty (MRU) and system robustness.

Main Results:

• Deterministic optimization reduced network losses and minimized total costs, outperforming Genetic Algorithms (GA) and Particle Swarm Optimization (PSO).
• Robust optimization determined system robustness levels under various uncertainty budgets, identifying sensitive parameters.
• Multi-objective robust optimization demonstrated a balance in uncertain parameter sensitivity, with constraints unmet beyond a 40% uncertainty budget.

Conclusions:

• The proposed robust optimization approach effectively enhances MG reliability against uncertainties.
• Multi-objective optimization provides a balanced trade-off between conflicting objectives and parameter sensitivities.
• The study confirms the superior capability of the FDA-based deterministic approach and highlights the importance of robust strategies for MGs.