Predictive energy management strategy of fuel cell bus based on comprehensive energy consumption and compound braking proportion distribution

To improve the energy efficiency of fuel cell buses and reduce the adverse effects on vehicle braking impact caused by the distribution of motor braking and mechanical braking, a predictive energy management strategy (EMS) for fuel cell buses with comprehensive energy consumption and compound braking ratio is proposed to weigh braking ride comfort and energy consumption. Firstly, the driver pattern recognition based on K-means is designed to improve the speed prediction module of BiLSTM (Bi-directional Long Short-Term Memory) to improve the speed prediction accuracy; then, the optimal braking distribution ratio and energy distribution strategy of fuel cell bus are optimized by designing LQR (linear quadratic regulator) controller, and the results are compared with the EMS based on CDCS (Charge depletion charge sustaining) and the EMS based on DP (Dynamic programming). The results show that this strategy can effectively reduce the energy consumption and braking impact degree while reducing the predicted RMSE (Root mean square value) by 9.4%. Compared to the regular braking ratio, the braking impact degree of this strategy is reduced from 3.02m/s3 to 2.72m/s3. By comparing the two benchmark EMSs, the performance of this strategy is close to that of the DP-based EMS. The equivalent hydrogen consumption of 100 km decreased from 7.02kg to 6.36kg.