Dynamic performance of liquid air energy storage system during discharge and grid connection with PSO-optimized PID control

Liquid air energy storage (LAES) is a promising solution for large-scale grid balancing. However, existing studies are largely limited to steady-state design, with insufficient evaluation of system dynamic characteristics under complex operating conditions. To address this gap, a dynamic simulation model of discharging process of LAES system is developed, and a controller parameter optimization strategy based on a parallel particle swarm optimization (PSO) algorithm is proposed within the system control framework. Simulation results show that the optimized control strategy effectively reduces high-frequency oscillations during startup and significantly reduces the rotor speed settling time from 48 s to 14 s. Under external thermal disturbances, the system demonstrates good robustness, restoring rated power within 88 s even under a severe reheating temperature drop of 100 K. Under grid frequency step disturbances, the system demonstrates a stable dynamic response, achieving a steady transition within 10 s. Overall, this study verifies the effectiveness and applicability of the developed dynamic model and its associated adaptive control system for LAES under varying operating conditions. The results demonstrate that the PSO-based PID parameter optimization method has significant potential to support stable grid operation and flexible dispatch in power systems with high renewable energy penetration. Furthermore, the findings provide a solid theoretical foundation and practical guidance for dynamic performance analysis, control strategy design, and actuator selection of hundred-megawatt-scale grid-connected LAES systems.