对于能源短缺和环境污染问题，发展以电动汽车及混合动力汽车为主的新能源汽车是一个有效的解决途径，储能系统则是这些汽车的关键组件。以锂离子电池作为储能系统的新能源汽车现已较多。但锂离子电池的耐久性仍需要进一步提高，其置换成本还很高。尤其是在低温情况下，电池的容量衰减速率加快，且电池的能量和功率特性变差，导致车辆的续驶里程缩短。因此，如何解决储能系统的高成本和短寿命问题，已成为汽车工业界和学术界共同关注的问题。电池/超级电容复合储能系统的出现为上述问题的解决提供了一个很好的方案，本文以电动客车为研究对象，建立其动力系统模型，并对复合储能系统的运行成本进行优化，最后设计了系统控制算法并完成了实验验证。首先，建立了客车的动力学模型以及复合储能系统模型，其中，复合储能系统模型包括超级电容的损耗模型、电池的损耗模型和电池的热模型。基于累积损耗理论，对阿伦尼乌斯（Arrhenius）衰减模型进行差分变形，考虑电池的放电深度、充放电倍率以及环境温度，建立了LiFePO4电池的动态容量衰减模型。利用实验数据对所建模型进行了标定和验证。其次，针对一种典型的半主动式储能系统，以客车的续驶里程为边界条件，确定了系统中电池的数量。利用动态规划算法，以复合储能系统在中国典型公交工况下的运行成本（包括电成本和电池衰减成本）为优化目标，对超级电容的数量以及复合储能系统的能量管理算法进行了联合优化，阐明了参数优化和能量管理算法优化间的相互影响关系。通过分析动态规划结果，找到了该复合储能系统可以在线使用的最优能量管理算法。之后针对复合储能系统低温使用时性能变差的问题，进行了系统自加热的仿真分析，阐明了系统自加热必要性同客车行驶里程、电池价格、加热效率以及环境温度之间的关系。基于上述工作，对四种不同构型的半主动式储能系统进行了优化和分析，比较了不同构型的储能系统的运行成本、购置成本、工作性能和控制简便程度。最后，基于主动式复合储能系统的五阶平均值模型设计了滑模控制器，对电池和超级电容的电流进行控制，设计了基于李亚普诺夫（Lyapunov）稳定性定律的总线电压控制器，通过仿真和实验验证了底层控制算法的有效性和鲁棒性。
The new energy vehicle (NEV), which mainly consists of electric vehicles (EVs) and hybrid electric vehicles (HEVs), is considered as an effective solution for the energy shortage and the environmental pollution. The energy storage system (ESS) plays a key role in those vehicles. Many NEVs use the lithium ion battery as the ESS. However, the durability of lithium battery needs to be further improved, and the cost of replacing the battery is high. Especially in low temperature environments, the battery degradation rate increases, and EVs suffer from significant driving range loss due to the resulting reduced energy and power capabilities of Li-ion batteries. Therefore, the research on solving the high cost and poor durability problems related to ESSs has gained consistent attention from both automobile industry and academia. The battery/supercapacitor hybrid energy storage system (HESS) is a good solution to solve the problem mentioned above. Taking the electric bus as research object, this dissertation focuses on the modeling of the vehicle powertrain, the operation cost optimization, and the control algorithm design and validation of HESS.Firstly, the vehicle dynamic model and the HESS model which includes the supercapacitor loss model, the battery loss model, and the battery thermal model, are established. Based on the cumulative damage theory, the dynamic degradation model of the LiFePO4 battery, which considers the influence of depth of discharge, charge/discharge rates, and ambient temperature, is proposed by using the differential form of the Arrhenius degradation model. The adopted HESS model is calibrated and verified by the experiment.Secondly, regarding a typical semi-active HESS, the battery amount is obtained according to the requirement of the electric bus minimal mileage. By using the Dynamic Programming (DP) approach, the supercapacitor amount and the EMS are optimized simultaneously along the typical China Bus Driving Cycle (CBDC), which minimizes the operation cost of the HESS (includes the electricity cost and the battery degradation cost). And the mutual influence between the parameter and the EMS optimization processes is explained. The optimal EMS for on-line uses is obtained by analyzing the DP result. Regarding the poor performance of HESS at subzero temperatures, the simulation of HESS self-heating is conducted, which reveals the relationship of the heating necessity with the bus driving range, the battery price, the heating efficiency, and the ambient temperature. Based on the above work, four semi-active HESS topologies are optimized and analyzed. The operation cost, the initial cost, the performance, and the control simplicity of different HESS topologies are compared. Finally, based on the 5th-order average model of a fully-active HESS, a sliding-mode controller is designed to control battery and supercapacitor currents. Meanwhile, a Lyapunov function-based controller is proposed to regulate the DC bus voltage. The effectiveness and robustness of the proposed control algorithm are validated by simulations and experiments.
