Dual-mode control and optimized power allocation for a flywheel-battery hybrid energy storage system

Abstract

Flywheel-lithium battery hybrid energy storage system (HESS) had become a pivot node for the widespread application of renewable energy, and charge–discharge control and power allocation strategies played a critical role in efficiency and lifetime of the whole system. Oriented to the dynamic characteristics of the flywheel energy storage system (FESS), a dual-mode regulation method was proposed to the charging and discharging process, where closed-loop linear active disturbance rejection control (LADRC) was addressed to develop speed and power control of the permanent magnet synchronous machine (PMSM) under basic runtime mechanism. In light of top-level power allocation, variational mode decomposition (VMD) was employed to perform adaptive high-fidelity decomposition of power signals, and non-dominated sorting genetic algorithm II (NSGA-II) was adopted to conduct multi-objective optimization of the crucial parameters of VMD simultaneously. Based on the VMD frequency spectrum, high-frequency power fluctuations were allocated to FESS and low-frequency components to battery energy storage system (BESS), which realized an optimized and coordinated control model of power allocation. To verify the proposed method, a numerical simulation by MATLAB/Simulink was adopted. The results indicate that compared with the traditional proportional-integral (PI) control, the proposed LADRC speed control reduced analysis time by 43.75% and the steady-state error by 60%, significantly enhancing the dynamic response and speed control accuracy of FESS. Compared with the power uniform distribution, the proposed optimal allocation method reduced the charge–discharge switching events of the BESS by 42.7% and 39.6% in the two cases, respectively, effectively achieving frequency-divided power optimization, significantly reducing the charge-discharge transitions, extending the system service life, and furthermore improving the comprehensive performance of HESS.