Multiobjective Model Predictive Control of Hydrogen Absorption in Solid-State Storage Optimized by Particle Swarm Algorithm

Abstract

Hydrogen is one of the most critical future fuels. To enhance hydrogen storage and transportation capabilities and bridge the gap with hydrogen utilization technologies, this study focuses on the hydrogen absorption process of solid-state hydrogen storage using LaNi₅ as the storage material. Model predictive control (MPC) was employed to optimize and regulate the thermal-pressure management of the container, investigating the regulatory methodology during the solid-state hydrogen storage reaction process and comparing it with proportional–integral-derivative (PID) control under specific operating conditions. By determining the mass source equation—which describes the hydrogen absorption and desorption reaction rates—based on the actual required flow rate at the terminal and performing backward deduction, and by coordinating three regulatory measures (valve opening, cooling power, and pressure regulation), the hydrogen reaction rate was consistently maintained within specified requirements to achieve optimal real-time optimization. Furthermore, a particle swarm optimization (PSO) algorithm was adopted to enhance the regulatory precision of MPC. Sensitivity analysis was conducted using the normalization of multiple reference indices to select the optimal prediction horizon and error coefficients for the objective function. Simulation results demonstrate that the reaction times under the set operating conditions are all within 225 s.