基于物质循环的氢燃料电池系统耦合机理及控制方法研究

It is a research hotspot of a highly efficient and long working lifetime proton electrolyte membrane (PEM) fuel cell system, which is an ideal power source for a future vehicle. A novel fuel cell system with a gas recycling mechanism can fully utilize water and heat that are generated by electro-chemical reactions inside the stack. Fresh hydrogen and air can be heated and humidified by recycled gases, and the outside humidifier is no longer required. By removing the humidifier, system efficiency, cold-start ability as well as system durability can be improved. However, sub-systems of such a fuel cell system are strongly coupled, and the control strategy becomes a challenging problem. This project will focus on the dynamic modeling and control of a fuel cell system with a gas recycling mechanism both for the cathode and anode. (1) A 0D+1D+1D multi-physical coupling model will be built up, and the full-order, reduced-order, analytical and error models will be deduced based on spatial discretization method, Hankel singular value, Adomian decomposition method and matrix polytope method. (2) A highly accurate states observation algorithm will be developed for inner states such as pressures of oxygen, hydrogen, nitrogen and vapor, temperature and humidity, as well as liquid water content in the gas diffusion layers of both the two sides. (3) A multi-objective robust adaptive control strategy will be proposed for four typical operation statuses, e.g. starting up, shutting down, diagnosis (typically for water flooding and membrane drying) and normal operation. Functions like intelligent start-stops, diagnosis, passive/active tolerant controls and self-humidified operation will be realized. By doing research on the three aspects, the mechanism of the novel fuel cell system will be explained, and the ultimate performances in efficiency and durability will be exploited. This project belongs to the multi-discipline domain of new energy power sources, control theory applications and electric-mechanic-thermal coupling systems. Research outcomes will provide theoretical guidance for the engineering design of fuel cell systems, which has important significance in promoting the development of fuel cell electric vehicles.

高效长寿命质子交换膜燃料电池系统，是未来交通运载工具的理想动力源，也是当前研究热点。采用氢气、空气循环结构的燃料电池系统可实现内部水分自平衡，有利于简化结构、提高效率、延长寿命，但系统耦合程度高、控制复杂。本项目拟针对此类燃料电池系统，(1)建立0D+1D+1D多物理场耦合机理模型，推导全阶、降阶、解析和误差模型，解释系统动态机理；(2)围绕内部气体（氧气、氢气、氮气、水蒸气等）的压强、温度、湿度，以及气体扩散层中的液态水含量等状态，研究高精度状态观测算法；(3)针对启动、工作、停机、故障等四个运行状态，研究多目标鲁棒自适应控制策略，实现智能启停、故障诊断、主/被动容错控制和自增湿运行等功能。通过上述工作，阐述系统工作机理，探究效率/寿命的极限性能。本项目属于新能源热-电耦合系统设计和控制交叉领域，研究结果将为燃料电池系统工程化设计提供理论指导，对促进行业发展具有重要意义。

高效长寿命质子交换膜燃料电池系统，是未来交通运载工具的理想动力源，也是行业热点。在本项目支持下，课题组针对具备氢气、空气双循环结构的新型燃料电池系统开展研究，包括如下方面：1）氢燃料电池降维动态机理模型及其验证；2）基于交流阻抗和外围传感器信息的状态观测；3）全工况氢-空-水-热鲁棒控制策略。通过该课题，建立了氢燃料电池降维动态机理模型（包括普通工况模型、非对称充排水动态模型、零下低温启动过程动态模型等），发明了多点气体采样系统，形成了燃料电池系统动态模型（误差在5%以内）。突破了燃料电池中高频交流阻抗高精度在线测试技术瓶颈，实现了中高频交流阻抗在线辨识，误差不超过2%（某些工况下最高可达到0.3%）。提出了基于交流阻抗和多传感器融合的燃料电池状态辨识算法，实现了空/氢侧气体压力、电堆温度、膜电极GDL液态水饱和度等内部状态变量的在线观测算法，稳态误差不超过5%。发明了燃料电池双循环系统构型，建立了双循环燃料电池系统氢-空-水-热-电协调控制体系，动态工况下流量控制误差降低一半(从5~10%到2~3%)，氢气压力波动降低一个数量级(从20~30kPa到2~3kPa)。针对北方极寒环境下的燃料电池冷启动问题，建立了一维冷启动多相转化机理模型，提出了低温冷启动最优加载策略、燃料电池最优停机吹扫策略与最小能耗启动策略，实现了-30℃环境2.5min快速启动。相关技术应用于电堆和燃料电池系统设计，电堆额定功率指标达到2.2kW/L，超过丰田首款燃料电池汽车Mirai 1.8kW/L；电堆测试寿命从600小时大幅提升到5000小时，台架加速寿命实验达到10000h以上，实车寿命预计超10000小时；系统净输出功率超过100kW。课题成果向北京亿华通公司转化，获得了全球首个100台级燃料电池商用车订单，完成公交、物流、有轨列车等多款车型开发。课题培养博士生6名，硕士生4名，博士后1名，工程师1名；发表SCI论文37篇、EI论文8篇、中文核心期刊论文1篇；两篇ESI高被引；一篇国际会议论文二等奖。申请/授权发明专利11项（包括1项美国发明专利），获得软件著作权4项。获得2019年“汽车工业科技进步一等奖”（负责人排第三），并获得日本丰田公司三年1500W的经费支持，用于继续开展面向燃料电池耐久性的系统研究。研究成果为在清华大学车辆学院建立燃料电池动力学科体系奠定了基础。

内容获取失败，请点击重试