V01.52 NO.3
Jun.2017
西南交通大学学报
UNIVERSITY
OF SOUTHWEST JIAOTONG
JOURNAL
第52卷第3期
2017年6月
文章编号:0258-2724(2017)03-0626-07
IX)I:10.3969/j.issn.0258-2724.2017.03.026
基于PNNPID的
车辆加速度控制器设计
梁
军1”,
赵彤阳1,
熊晓夏1,
宁2
陈
朱
张婉婉1,
(1.江苏大学汽车工程研究院,江苏镇江212013;2.静冈理工科大学机械系,静冈袋井437-0032)
龙1,
摘要:针对传统的自适应巡航(adaptive cruise control,ACC)加速度控制算法对系统变化辨识速度慢、动态性
neural network
proportion integration differentiation)的加
differentiation)对误差直接反馈
proportion integration
能差、调整缓慢的问题,提出了一种基于PNNPID(parallel
neural network
速度控制算法.通过分析传统SNNPID(serial
的不足,应用神经网络自学习功能,开发了基于并行控制原理的车辆加速度控制器;考虑车辆平顺性,设计了符
合驾驶行为的滤波限幅方案.实验结果表明,相比基于SNNPID的加速度控制器,基于PNNPID的加速度控制器
最大偏差控制在± O.25 m/s2以内,并具有平均误差小、调整时闭短和瞬态特性良好的特点,滤波限幅方案有效
提高了驾驶的舒适性.
关键词:自适应巡航;加速度控制器;动态性能;神经网络;PID
中圈分类号:U471.1
文献标志码:A
of Vehicle Acceleration Controller Based on
Design
PID
Parallel Neural Network
HANG
Junl”,ZHAO
Xiaoxial,
Tongyan91,XIONG
ZHANG
Wanwanl,CHENLon91,ZHU Nin92
of
Research
(1.Automotive Engineering
Institute,Jiangsu University,Zhenjiang 212013,China;2.Department
Mechanical
Institute of Science and
Engineering,Shizuoka
Technology,Shizuoka437-0032,Japan)
acceleration control
based on the
neural network
parallel
Abstract:An
PID(PNNPID)was
algorithm
to
the
problems
of slow
proposed
slow
conquer
system change identification,poor dynamic performance,and
in traditional acceleration control
for
cruise
control(ACC).
adjustment
algorithms
of the traditional serial neural network
shortcomings
adaptive
of the
direct
PID(SNNPID)for
Through analysis
for
is based on the
control
is
feedback error.all acceleration controller
vehicles that
theory
parallel
the
into account
the
function of neural network.In
scheme of filter is
developed using
self-learning
addition,taking
vehicle
to fit
behavior.
driving
ride
comfort,all amplitude limiting
designed
on
can function within
SNNPID
a
results show that the acceleration controller based
Experimental
maximum deviation of± 0.25
is more accurate than the controller based Oil PNNPID.This
m/s.and
收疆日期:2015· 12-30
基金项目:国家自然科学基金资助项目(61573171,51108209,61203244);江苏高校优势学科建设工程资助项目(PAPD);江苏省自然
科学基金资助项目(BK20140570);江苏省六大人才高峰项目(DZXX-048);中国博士后科学基金资助项目
(2016M600375);江苏省研究生教改课题(101)
作者简介:梁军(1976一),男,教授,研究方向为智能交通理论及应用,电话:0511-88782845,E—mail:liangjun@ujs.edu.cn
引文格式:梁军,赵彤阳,熊晓夏,等.基于PNNPID的车辆加速度控制器设计[J],西南交通大学学报,2017,52(3):626.632.
of vehicle acceleration controller based on
a1.Design
neural network
LIANG
Jun,ZHAOTongyang,XIONG Xiaoxia,et
parallel
PID[J].Journal ofSouthwest
Jiaotong University,2017,52(3):626&32.
万方数据
全部评论(0)