混合动力车辆综合了电动汽车经济性高及燃油汽车续航时间长的优势逐渐成为研究热点。随着技术的不断进步，未来混动车辆“源网荷储”互动将进一步加深。但是目前研究大部分集中于设备层，没有在系统层对各可控资源进行协同管理，提升指标较为片面，应对未来多样化负荷需求能力较弱，能量利用效率不足，亟需系统层面的能量管理系统进行统一协调调度。本文提出了一种协调发动机、动力电池、超级电容和动能回收的混动车辆能量管理系统调度控制架构，通过分钟级、秒级时间尺度下的双层迭代求解最终确定各个子设备功率输出，以应对不同时间尺度下的负荷需求。建模过程中，分钟级优化调度在负荷预测的前提下，利用考虑车辆燃油经济性和启停状态的发动机和动力电池协调长时间尺度下的功率需求，削峰填谷，并在电池热安全约束下两阶段迭代求解；秒级优化调度在更精确负荷预测下，多时段衔接约束求解超级电容和车辆发动机秒级功率响应；在秒级时间尺度负荷极端波动，模型无法求解时，启动紧急控制，更新边界条件，区分确定性负荷条件和不确定性负荷条件充分发挥各可控资源调节能力，最终形成各设备功率控制参考值。并利用全球统一轻型车辆测试循环工况和仿真模型对上述架构进行案例验证，证明了所提架构的实用性和有效性。本文通过上述控制架构充分利用车辆特性各异的各可控资源。在分钟级优化发动机燃油消耗率并利用动力电池进行负荷转移，保证动力电池热稳定的同时提高了车辆经济性。在秒级利用超级电容响应负荷功率需求和动能回收，注重负荷实时功率满足的同时优化发动机燃油经济性，进一步提高车辆续航水平。在前述模型中，本文还考虑动力电池和超级电容荷电状态与功率关系并模拟了其荷电状态变化。对于负荷的秒级概率特征，利用偏态分布对负荷概率进行拟合，提出了紧急控制条件下负荷最大化供电的概率调度模型。
Hybrid electric vehicles integrate the advantages of high economy of electric vehicles and long endurance of fuel vehicles and gradually become research hotspots. With the continuous progress of technology, the interaction of "source network load storage" of hybrid electric vehicles in the future will be further deepened. However, most of the current research focuses on the equipment level, there is no coordinated management of controllable resources at the system level, the improvement index is relatively one-sided, the ability to cope with the future diversified load demand is weak, and the energy utilization efficiency is insufficient, so it is urgent for the energy management system at the system level to carry out unified coordination scheduling.A hybrid electric vehicle energy management system control scheduling architecture is proposed in this paper, which coordinates engine, power battery, supercapacitor and kinetic energy recovery. The power output of each sub-facility is finally determined by two-layer iterative solution at minute and second time scales to meet the load demand at different time scales. In the modeling process, the minute-level optimal scheduling uses the engine and power battery considering fuel economy and start-stop state to coordinate the power demand on a long time scale under the premise of load prediction, clip peaks and fill valleys, and solves the problem iteratively in two stages under battery thermal safety constraints; second-level optimal scheduling solves supercapacitor and vehicle engine second-level power response under multi-period convergence constraint when load prediction is more accurate; when the load fluctuates extremely in the second time scale and the model cannot be solved, the emergency control is started to update the boundary conditions, distinguish the deterministic load conditions and the uncertain load conditions, give full play to the adjustment ability of each controllable resource, and form the power control reference value of each facility. Finally, the proposed architecture is verified by a case study of WLTC test cycle and simulation model, which proves the practicability and effectiveness of the proposed architecture.In this paper, the controllable resources with different characteristics of vehicles are fully utilized through the above control architecture. The fuel consumption rate of the engine is optimized at the minute level and the load transfer is carried out by using the power battery to ensure the thermal stability of the power battery and improve the vehicle economy. At the second level, the super capacitor is used to respond to load power demand and kinetic energy recovery, focusing on real-time load power satisfaction while optimizing engine fuel economy, further improving vehicle endurance level.In the above model, the relationship between SOC and power of power battery and super capacitor is also considered and the change of SOC is simulated. For the second level probability characteristics of load, the load probability is fitted by skewness distribution, and a probabilistic dispatching model for load maximization under emergency control conditions is proposed.
