摘要
为了改善车辆动力性、提高燃油经济性、减小车辆的冲击度以及提高乘坐的舒适性,将T-S模糊神经网络应用到自动换挡上,以车速和油门开度作为T-S模糊神经网络的2个输入参数。首先选取训练样本,通过对样本进行训练得到准确的T-S模糊神经网络控制模型,然后在matlab/simulink中搭建仿真模型。在先加速后减速的工况下仿真,仿真结果表明车辆换挡点误差在2%以内,证明了T-S模糊神经网络应用在自动换挡上的可行性,以及较强的鲁棒性。
In order to improve vehicle power performance and fuel economy, to reduce vehicle impact and improve ride comfort, T-S fuzzy neural network is applied to automatic shift. Vehicle speed and throttle opening are taken as two input parameters of T-S fuzzy neural network. First, samples are selected and trained to obtain the accurate T-S fuzzy neural network control model. Then the vehicle model is built in matlab/simulink. The simulation results show that the shift point error is less than 2%, which proves the feasibility and robustness of T-S fuzzy neural network applied in automatic shift.
引文
[1] Fischer,Robert.The automotive transmission book[M].Cham:Springer International Publishing AG,2016.
[2] 陈桥,姚广涛,张英锋,等.自动变速器换挡品质评价指标体系研究现状[J].装备制造技术,2017(05):50-52.
[3] 荆崇波,苑士华,郭晓林.双离合器自动变速器及其应用前景分析[J].机械传动,2005(03):56-58.
[4] 吴光强,孙贤安.汽车自动变速器发展综述[J].同济大学学报,2010,38(10):1478-1483.
[5] 赵丁选,崔功杰.基于改进BP网络的工程车辆自动换挡策略[J].华南理工大学学报,2008(10):67-70.
[6] 李平康,金涛涛,李蓓.基于模糊规则递推的车辆自动换挡策略仿真研究[J].北京交通大学学报,2008(01):88-92.
[7] Yi Jun,Wang Xuelin,Hu Yujin.Fuzzy control and simulation on automatic transmission of tracked vehicle in complicated driving conditions[C].Vehicular Electronics and Safety,IEEE International Conference,2006:259-264.
[8] 陈清洪,秦大同,叶心.AMT汽车动态模糊神经网络三参数换挡策略研究[J].汽车工程,2010,32(06):505-509.
[9] 孟凡婧,金辉.自动变速车辆下坡换挡策略研究[J].汽车工程,2018,40(02):192-199.
[10] 崔功杰.工程车辆三参数最佳换挡规律及控制方法研究[D].长春:吉林大学,2009.
[11] 肖建,赵涛.T-S模糊控制综述与展望[J].西南交通大学学报,2016,51(03):462-474.
[12] 袁宇浩,张广明.T-S模糊广义系统研究综述[J].自动化学报,2010,36(07):901-911.
[13] 钱超.基于T-S模糊控制的纯电动汽车换档规律研究[D].合肥:合肥工业大学,2017.
[14] 刘振军,胡建军,李光辉,等.模糊神经网络的自动变速汽车换挡规律分析[J].重庆大学学报,2009,32(08):897-90.