.png){kind=icon}
.png){kind=icon}
Return
A Cost-Effective Real-Time Energy Management Strategy for Fuel Cell Electric Vehicles Combining Dynamic Programming and Model Predictive Control
.png){kind=icon}
PRE
AI
R
D
Y
DOI：10.1109/tvt.2026.3665713.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Fuel cell electric vehicles (FCEVs) encounter difficulties in minimizing running costs and prolonging fuel cell lifespan, and current energy management strategies fail to reconcile real-time performance with global optimization. This study proposed a framework integrating model predictive control (MPC) and dynamic programming (DP) to reduce the running cost of the FCEVs' powertrain system. A bidirectional long short-term memory (BiLSTM) network optimized via the dung beetle optimizer (DBO) algorithm enabled high-accuracy velocity forecasting, and a multi-objective cost function incorporating equivalent hydrogen cost, lithium battery pack loss, and fuel cell loss was established. Leveraging the state of charge (SOC) of the lithium battery, the DP algorithm determined the best power distribution within the control horizon, and energy utilization was optimized through the rolling optimization of MPC. The findings indicate that the suggested DP-MPC strategy markedly diminishes fuel cell power fluctuations and loss costs, lowering the overall running cost of the FCEVs powertrain system throughout four driving cycles: the Worldwide Harmonized Light Vehicles Test Procedure (WLTP), the Supplemental Federal Test Procedure (US06), the Urban Dynamometer Driving Schedule (UDDS), and the China Light-Duty Vehicle Test Cycle (CLTC). Compared to DP, sequential quadratic programming MPC (MPC-SQP), Pontryagin's Minimum Principle minimizing equivalent hydrogen consumption (PMP-H2-min), DP minimizing equivalent hydrogen consumption (DP-H2-min), and rule-based strategies, DP-MPC achieves average fuel cell loss reductions of 26.4%, 22.0%, 67.0%, 60.5%, and 71.9%, respectively, and reduces the total running cost of the FCEV by 10.6%, 21.2%, 16.9%, and 41.7% compared to MPC-SQP, PMP-H2-min, DP-H2-min, and rule-based strategies. The numerical and hardware-in-the-loop (HIL) experimental results show that the proposed strategy offers excellent direction for decreasing FCEVs’ running costs and alleviating fuel cells’ degradation.
Keywords:
Fuel cell electric vehicles
speed prediction
dynamic programming
model predictive control
running cost
Journal
IF:
7.1
Papers:
1.7W
Citations:
6.6W
Organization
Cited Papers