基于路况预测的混合动力汽车智能能量管理策略研究

混合动力汽车因兼具传统汽车与纯电动汽车的优势而广受关注，但其能量管理策略运行效果受行驶路况影响大，导致其整车能效水平远未达到最佳。因此，拟首先基于ITS、车联网及云计算平台实现路况大数据的准确获取与处理分析，进而基于深度增强学习研究考虑“人-车-路-环境”交互影响的的复杂路况系统智能预测方法；其次，建立动力电池电-热-老化多领域耦合模型并基于强跟踪自适应粒子滤波实现电池多状态联合估计；再次，基于滚动优化思想、鲁棒预演控制理论及云计算等新理论新方法，研究综合利用路况预测信息的混合动力汽车在线滚动优化智能能量管理策略。最后，设计基于云计算的硬件/车辆在环试验平台，验证提出的新理论新方法。本课题涉及控制理论、车辆工程、交通、大数据等多学科交叉领域，对促进相关学科的基础理论研究和应用、发展我国混合动力汽车及其相关产业、解决环境能源危机具有重大的理论价值和现实意义。

Having the advantages of both traditional and pure electric vehicle，hybrid eletric vehicle receives universal concerns. However, the energy efficiency of the hybrid electric vehicle is far from the optimal level because the effect of energy management strategy is affected by the traffic conditions. Therefore, firstly, based on ITS, vehicle networking, and based on depth-enhanced learning, an intelligent prediction method for complex road conditions considering the interaction of “human-vehicle-road” will be explored; Secondly, build the electric - heat - aging multi - field coupling model of the power battery and realize the battery multi - state robust joint estimation based on the strong tracking adaptive particle filter. Thirdly, study the intelligent energy management strategy for hybrid electric vehicles by using the road condition prediction and based on the theory of rolling optimization, robust preview control theory and cloud computing technology. Finally, design a cloud computing-based hardware/vehicle - in - the - loop test platform and verify the validity of the new method and theory. This topic involves interdisciplinary areas of control theory, vehicle engineering, traffic information and big data etc. and has great theoretical value and practical significance in promoting the basic theory research and application of related disciplines, promoting the development of hybrid electric vehicle and its related industries of our country, and solving the environmental energy crisis.