摘要
为保障辅助驾驶模式下汽车具备良好的制动减速性能,提出了一种基于模糊神经网络的体现驾驶员制动特点的辅助制动系统,该系统以两车的相对速度和车间距离为网络输入,以汽车制动力为网络输出。利用驾驶员在不同车况下的合理而充足的制动数据对网络参数进行离线训练,优化网络权值,建立了体现驾驶员制动特点的模糊神经网络控制器模型。进而在Matlab/Simulink中建立汽车动力学模型与基于滑移率的ABS模型进行仿真,仿真结果表明汽车辅助制动系统能很好地起到减速制动效果。
For ensuring good braking performance of vehicle under auxiliary driving mode,a fuzzy neural network（FNN）-based auxiliary braking system,embodying drivers＇ behavior of brake application is proposed,which takes the speed difference and distance between two vehicles as network input with braking force as network output.Ample reasonable braking data of drivers in different driving conditions are used to perform offline training on network parameters with network weights optimized,and thus a controller model for FNN,reflecting drivers＇ behavior in applying brake is built.Further more,the vehicle dynamics model and slip rate-based ABS model are developed in Matlab/Simulink and a simulation proceeds with the results showing that the system exhibits satisfactory braking effectiveness.
出处
《汽车工程》
EI
CSCD
北大核心
2010年第12期1071-1076,1101,共7页
Automotive Engineering
基金
湖南大学汽车车身先进设计制造国家重点实验室开放基金(30815005)
教育部长江学者与创新团队发展计划(531105050037)
国家人事部留学人员科技活动择优项目(2007)
长沙市科技计划重点项目(K0803104-11)资助
