Optimizing multi-energy storage in hybrid energy hubs: A stochastic approach using an improved artificial hummingbird algorithm for renewable integration and cost reduction

Combining different energy sources in hybrid energy systems has attracted significant attention due to their exceptional efficiency, adaptability, and eco-friendliness. Effective operation of these systems relies heavily on optimal planning and management. This research introduces a thorough framework for the efficient planning and energy management of a hybrid energy hub (EH) that includes electric, thermal, and cold storage. The main goal is to minimize the costs associated with operation and storage utilization by framing the issue as a stochastic optimization problem. An advanced stochastic methodology is utilized to tackle uncertainties arising from variable energy sources and demand, using Monte Carlo Simulation combined with K-means clustering for scenario generation and reduction. This method integrates the Improved Artificial Hummingbird Algorithm (IAHA) with a population classification strategy, Monte Carlo simulation-based probabilistic analysis, and K-means clustering. The effectiveness of the proposed approach is confirmed through simulations conducted on a standard EH under various scenarios. These scenarios include energy management without and with multi-energy storage and uncertainty consideration. The results indicate that integrating storage into EH planning and management using IAHA leads to an 11.84 % reduction in operational costs. Additionally, solving the stochastic problem through Monte Carlo simulation and K-means clustering results in a significant 36.30 % reduction in operational costs, allowing for more precise decisions regarding EH effectiveness. This study emphasizes the IAHA's effectiveness in optimizing EH energy management, showcasing its superiority over traditional optimization algorithms such as AHA, particle swarm optimization (PSO), gray wolf optimizer (GWO), manta ray foraging optimizer (MRFO), and Kepler optimization algorithm (KOA).