基于Stackelberg博弈的燃料电池混合动力汽车跟车能量管理  

Car-following Energy Management of Fuel Cell Hybrid Electric Vehicles based on Stackelberg Game

作　　者：付主木 朱龙龙[1] 陶发展 李梦杨 FU Zhumu;ZHU Longlong;TAO Fazhan;LI Mengyang(Information Engineering School,Henan University of Science&Technology,Luoyang 471023,China;Henan Key Laboratory of Robot and Intelligent Systems,Henan University of Science&Technology,Luoyang 471023,China;College of Physics&Electronic Information,Luoyang Normal University,Luoyang 471934,China)

机构地区：[1]河南科技大学信息工程学院,河南洛阳471023 [2]河南科技大学河南省机器人与智能系统重点实验室,河南洛阳471023 [3]洛阳师范学院物理与电子信息学院,河南洛阳471934

出　　处：《河南科技大学学报（自然科学版）》2024年第4期1-9,M0002,共10页Journal of Henan University of Science And Technology:Natural Science

基　　金：国家自然科学基金项目(62301212,62371182);河南省高校科技创新人才计划项目(23HASTIT021);龙门实验室重大科技项目(231100220300);河南省科技研发计划联合基金项目(222103810036);河南省重点研发与推广专项科技攻关项目(222102240009)。

摘　　要：跟车场景下燃料电池混合动力汽车(FCHEV)的速度与能量管理协同优化是实现车辆节能的重要有效手段,针对现有策略中双能量源退化与能耗耦合关系不明,且难以兼顾全局优化与实时性能的问题,提出一种基于Stackelberg博弈的FCHEV跟车能量管理策略。首先,建立了燃料电池/锂电池的能耗与性能退化模型,并纳入到统一量纲的整车综合使用成本函数中;其次,提出了基于分层解耦的跟车能量管理策略,实现跟车速度与功率分配的解耦控制;最后,综合考虑跟车安全性、舒适性、燃料经济性和能源耐久性,建立跟车控制层与能量管理层对应的双层规划模型,并基于Stackelberg博弈思想设计了双层差分遗传算法对策略核心参数进行离线优化。仿真和实验结果表明:相较于模型预测控制方法,该方法可降低平均车间距误差37.7%、平均冲击度2.4%、等效氢气消耗9.3%和能源退化成本13.9%,实现了优化性能与实时性的兼顾。The collaborative optimization of speed and energy management for fuel cell hybrid electric vehicles(FCHEVs)in car-following scenario is an important and effective means to achieve vehicle energy conservation.In view of the unclear coupling relationship between dual energy source degradation and energy consumption in existing strategies,and the difficulty of considering both global optimization and real-time performance,this paper proposes a Stackelberg game based FCHEV following energy management strategy.Firstly,models for energy consumption and performance degradation of fuel cells/lithium batteries are established and incorporated into a unified dimensional comprehensive vehicle cost function.Secondly,a hierarchical decoupling based energy management strategy for vehicle following is proposed to achieve decoupling control of vehicle following speed and power distribution.Thirdly,considering safety,comfort,fuel economy,and energy durability,a bi-level programming model corresponding to the car-following control layer and energy management layer is established.Based on the Stackelberg game theory,a bi-level differential genetic algorithm is designed to offline optimize the core parameters of the strategy.Finally,simulation and experimental results show that compared to the model predictive control method,this method can reduce average vehicle spacing error by 37.7%,average impact by 2.4%,equivalent hydrogen consumption by 9.3%,and energy degradation cost by 13.9%,achieving a balance between optimized performance and real-time performance.

关 键 词：燃料电池混合动力汽车 跟车能量管理 双层规划 STACKELBERG博弈 双层差分遗传算法 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]