Arrhenius-Based Capacity and Dispatch Optimization of a Wind–Photovoltaic–Storage Microgrid Using Task-Preserving, Degradation-Aware Refinement
Year: 2026
Type: preprint
Abstract: Feasibility repair can change the state-of-charge (SOC) trajectory used to evaluate battery degradation. We develop a cost–carbon–degradation model for a grid-connected wind–photovoltaic–storage microgrid. After correcting each schedule to meet a common daily terminal SOC, a degradation-aware refinement shifts charging and discharging between admissible hours while preserving their respective daily energy totals. An improved nondominated sorting genetic algorithm II (NSGA-II) combining differential evolution, adaptive variation and structured local search uses 24,000 objective evaluations in each of 12 independent runs. Across the 12 runs, the improved NSGA-II achieved a mean inverted generational distance (IGD) of 0.02840 ± 0.00168 and a normalized hypervolume (HV) of 0.84665 ± 0.00444. The corresponding mean values for standard NSGA-II were 0.08489 and 0.72758. The representative solution from the run closest to the median IGD is rated at 40.49 MW/114.97 MWh. Across 12 candidate pairs with matched ratings and daily charge and discharge totals, refinement reduces modeled first-year capacity loss by 10.70% ± 1.11% on average. Fixed-trajectory sensitivity across 17 degradation-model settings gives mean reductions in modeled first-year loss of 0.39% to 16.19%; two candidates instead show increased loss in the anchored linear case. These results show that the degradation benefit of temporal redistribution depends on the assumed aging model and that feasibility-repair choices can affect degradation assessment.
Source: Research Square
Institution: Shandong Agricultural University
Language: English
Subfield: Control and Systems Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7225988705
