氢燃料电池汽车不同制氢方案的全生命周期评价及情景模拟研究

为有效地提高插电式燃料电池汽车的经济性，实现燃料电池和动力电池的功率最优分配，考虑到行驶工况、电池荷电状态（State of charge, SOC）、等效因子与氢气消耗之间的密切联系，制定融合工况预测的里程自适应等效氢耗最小策略. 通过基于误差反向传播的神经网络来实现未来短期车速的预测，分析未来车辆需求功率变化，同时借助全球定位系统规划一条通往目的地的路径，智能交通系统便可获取整个行程的交通流量信息，利用行驶里程和SOC实时动态修正等效消耗最小策略中的等效因子，实现能量管理策略的自适应性. 基于MATLAB/Simulink软件，搭建整车仿真模型与传统的能量管理策略进行仿真对比验证. 仿真结果表明，采用基于神经网络的工况预测算法能够较好地预测未来短期工况，其预测精度相较于马尔可夫方法提高12.5%，所提出的能量管理策略在城市道路循环工况(UDDS)下的氢气消耗比电量消耗维持(CD/CS)策略下降55.6%. 硬件在环试验表明，在市郊循环工况 (EUDC)下的氢气消耗比CD/CS策略下降26.8%，仿真验证结果表明了所提出的策略相比于CD/CS策略在氢气消耗方面的优越性能，并通过硬件在环实验验证了所提策略的有效性.

Life-cycle assessment and scenario simulation of four hydrogen production schemes for hydrogen fuel cell vehicles

The environment pollution and petroleum problems, which are increasingly becoming serious, have caused the vehicle industry to transition into a low-carbon and energy-saving industry. During processes, plug-in fuel-cell electric vehicles (PFCEVs) play an important role due to their advantages of rapid fueling, high energy density and efficiency, low operating temperature, and zero onboard emissions. PFCEVs use high-capacity rechargeable batteries to avoid working in low-efficiency areas. However, a robust energy management strategy that can achieve reliable energy distribution by regulating the output power of the fuel cell and battery within the hybrid powertrain merits further investigation. Considering the close relationship between the driving cycle, state of charge (SOC), equivalent factor, and hydrogen consumption, a trip distance adaptive equivalent consumption minimum strategy integrating driving cycle prediction is proposed. A backpropagation-based neural network is used to predict short-term vehicle velocity and analyze future changes in vehicle demand power. Planning a path to the destination with the help of the global positioning system, the intelligent transportation system can also obtain traffic flow information for the entire trip. The equivalent factor is dynamically corrected in real time using the driving distance and SOC to realize the adaptability of the energy management strategy. Finally, the velocity prediction sequence is combined with the objective function. The sequential quadratic programming algorithm is used to optimize the equivalent hydrogen consumption of the objective function and to obtain the distributed power of the fuel cell and battery. The vehicle simulation model is built and compared with a traditional energy management strategy based on MATLAB/Simulink software. The simulation results show that the driving cycle prediction algorithm based on the backpropagation-based neural network predicts future short-term conditions better, with a 12.5% higher accuracy than the Markov method. The proposed energy management strategy allows the fuel cell to operate in high-efficiency areas. The hydrogen consumption is 55.6% less than that of the CD/CS strategy under the UDDS cycle. The hardware in the loop experiment verifies a hydrogen consumption that is 26.8% less than that of the CD/CS strategy under the EUDC cycle. The numerical validation results demonstrate the superior performance of the proposed strategy in terms of hydrogen consumption over the CD/CS strategy. The effectiveness of the proposed strategy is validated by hardware during the loop experiment.