Sensitivity-Driven optimization analysis for renewable resource allocation and energy distribution systems

Abstract

Integration of renewable energy resources into modern-day energy systems poses numerous problems associated with uncertainty, resource allocation, and network stability. Existing optimization methods tend to consider only nominal operations and fail to ensure the robustness of the network under adverse scenarios with regard to variations in renewables, changes in demand, deteriorating network performance, and other factors. This paper presents the Graph-Sensitivity-Driven Minimax Optimization Framework, which allows robust renewable resource allocation and management of energy distribution. The presented method considers the renewable energy distribution network as a weighted graph, where generation nodes, storage nodes, substations, and demand nodes are considered nodes, while paths for energy transfer are considered edges. Sensitivity analysis of the network nodes and edges is conducted to find out the critical components of the network, whose deterioration could have a strong impact on overall network performance. The designed GSDOA framework examines uncertain conditions and optimizes the allocation of renewable generation and storage, as well as the decision regarding transmission of power, subject to certain operational constraints. Results of the experimental evaluation using various network distributions and renewable energy sources reveal that the designed framework exhibits better robustness than the deterministic optimization, cost optimization, conventional minimax robust optimization, and graph-insensitive robust optimization frameworks. The designed framework obtains a result where 95.4% renewable energy is used, energy loss is at 5.72%, connectivity loss is at 6.1%, and loss of worst case is 7.8%.