Abstract
This paper presents a new stochastic-intelligent framework for sizing and energy management of a hybrid renewable energy system consisting of photovoltaic (PV), wind turbine, and hydrogen energy storage-based fuel cells (PV/Wind/FC). The framework incorporates a cloud model to address uncertainties in renewable generation and system load, with aim of the cost of energy (COE) while satisfying the loss of energy probability (LOEP). An improved Walrus Optimizer (IWO) with a piecewise linear chaotic map is applied to determine the optimal system component sizes. The model's effectiveness is evaluated through deterministic and stochastic scenarios using real meteorological data from Beijing, Guangzhou, Kashi, and Xining, China. The deterministic results clear that the PV/Wind/FC system outperforms other configurations, achieving the lowest COE and LOEP. The COE values for Beijing, Guangzhou, Kashi, and Xining are 0.260, 0.202, 0.246, and 0.217 $/kWh, respectively. The IWO algorithm demonstrates superior performance compared to traditional methods such as WO, PSO, MRFO, and GWO in terms of COE, reliability, convergence speed, and stability. In the stochastic approach based on cloud model, the COE increases by 13.84%, 14.85%, 10.97%, and 15.66% for the respective regions, highlighting the impact of renewable generation and system demand uncertainties. Additionally, the cloud model findings demonstrate how uncertainty distributions impact the system's operation, with the variation in cloud model droplets on both sides of the expected value reflecting the effects of renewable generation and demand uncertainties. This provides a more comprehensive and reliable framework for HRES design under uncertain conditions compared to the deterministic model.