摘要
为提高插电式混合动力汽车的燃油经济性,根据城市循环工况的特点选定了4种典型的城市工况,采用学习向量量化(LVQ)神经网络识别车辆运行实时工况,并在MATLAB/Simulink平台制定了一种基于工况识别的整车控制策略.基于实例车型,在Cruise软件中建立了整车仿真模型,并在城市工况下进行仿真.仿真结果表明:所建立的控制策略能够有效识别工况信息;能够以此进行相应工作模式的切换和合理的转矩分配,且相对于传统汽车燃油经济性有明显的提高.从而验证了该控制策略的合理性和有效性.
To improve fuel economy of plug-in hybrid vehicles,with choosing four typical city conditions according to characteristics of city driving cycle,by using LVQ neural network to identify vehicle operating conditions, a control strategy was developed based on vehicle condition recognition on MATLAB/Simulink platform. Vehicle model was established in Cruise based on a real example model,and the model was simulated under representative working condition of city. The results show that the control strategy can effectively identify the working condition information,and switch the corresponding working mode and distribute torque reasonably. Fuel economy has been improved significantly compared with traditional vehicle,which verifies the rationality and effectiveness of control strategy.
作者
尹安东
姜涛
YINAn-dong;JIANGTao(School of Automobile and Traffic Engneemng,Hefei Univemity of Technology,Hefei 2300;Nation and Local Union Research Center for Automotive Technology ＆ EquiHefei 230009,China)
出处
《车辆与动力技术》
2018年第2期1-6,共6页
Vehicle & Power Technology
基金
国家科技支撑计划项目(2013BAG08B01)
国家新能源汽车产业技术创新工程资助项目(财建[2012]1095号)
关键词
插电式混合动力汽车
学习向量量化神经网络
工况识别
控制策略
燃油经济性
plug - in hybrid electric vehicle
learning vector quantization neural network
vehicle condition identification
control strategy
fuel economy
