The article discusses the significance of off-grid renewable energy hydrogen production in enhancing renewable energy utilization and power system stability. It addresses the challenges posed by the stochastic fluctuations of wind and solar power to electrolyzer reliability. The paper introduces a two-stage power optimization method that combines rule-based allocation with algorithmic optimization to improve the efficiency of wind-solar hydrogen production systems, taking into account reserved energy storage. In the first stage, the method determines hydrogen production power and hybrid energy storage systems (HESS) initial allocation based on real-time electrolyzer conditions and reserved energy. The second stage utilizes an improved multi-objective particle swarm optimization (IMOPSO) algorithm to optimize HESS power allocation, aiming to reduce unit hydrogen production cost and average battery charge-discharge depth. The proposed method aims to enhance hydrogen production stability, increase HESS supply capacity, reduce renewable curtailment rates, and lower average production costs. Case studies presented in the article highlight the effectiveness of the new approach compared to three conventional rule-based power allocation methods. Overall, the research contributes to the optimization of off-grid wind-solar hydrogen production systems, offering a promising solution to enhance energy sustainability and address challenges associated with power fluctuations in renewable energy production.