Optimal flexible power allocation energy management strategy for hybrid energy storage system with genetic algorithm based model predictive control

This paper proposes an optimal flexible power allocation-based energy management system (EMS) for hybrid energy storage systems (HESS) in electric vehicles (EVs). The main advantage of the proposed EMS is its use of a genetic algorithm-based model predictive control (GA-MPC) with optimal tuning of weights and constraints, which fully extends and balances ultracapacitors (UC) utilization over the actual driving cycle. Additionally, the GA-MPC gently isolates the battery from direct recharging, further extending the battery's lifetime. For the prediction aspect, a hybrid ARIMA-LSTM predictor is developed for real time prediction of more accurate power demand for the EMS optimization. Meanwhile, a MPC with a quadratic programming algorithm is designed to optimize the power allocation problem. Innovatively, a rule-based strategy and the GA are developed for constraint and weight regulation, gently preventing direct battery recharging and ensuring the UC has the highest priority in the dynamic power allocation process. The proposed approach is validated using a downscaled HESS experimental platform. Simulation results indicate that the UC utilization ratio is improved by 25.3 %, and the battery lifetime is prolonged by 13 % compared to conventional MPC, without altering the existing HESS topology. • A novel GA-MPC strategy is introduced to optimize power allocation and UC utilization, achieving an flexible power allocation. • A hybrid ARIMA-LSTM model is proposed to improve the accuracy of power demand prediction. • A rule-based method is employed for model selection, constraint regulation, and triggering the GA for weight optimization. • A scaled-down hardware-in-the-loop verification platform is established to validate the effectiveness of the proposed strategy.