基于驾驶意图多步预测的智能网联HEV等效排放最小控制策略  

Equivalent Emission Minimization Strategy of Intelligent Connected HEV Based on Multi-step Prediction of Driving Intention

作　　者：王跃飞[1,2] 王志 孙睿 王超 肖锴[1] 潘斌 WANG Yuefei;WANG Zhi;SUN Rui;WANG Chao;XIAO Kai;PAN Bin(School of Mechanical Engineering,Hefei University of Technology,Hefei 230009;Anhui Jianghuai Automobile Group Corp.,Ltd.,Hefei 230022)

机构地区：[1]合肥工业大学机械工程学院,合肥230009 [2]安徽江淮汽车集团股份有限公司,合肥230022

出　　处：《机械工程学报》2023年第18期271-282,共12页Journal of Mechanical Engineering

基　　金：国家自然科学基金(61202096);安徽省重点研究与开发计划(202104a05020018)资助项目。

摘　　要：车辆“碳达标”与“碳中和”是实现全社会碳达标的重要组成部分。为降低智能网联混合动力汽车(Intelligent connected HEV,ICHEV)总排放量,提出一种基于驾驶意图多步预测(Multi-step prediction of driving intention,MPDI)的ICHEV等效排放最小(Equivalent emissions minimization strategy,EEMS)能量在线管理策略。首先,给出车联网V2X下驾驶意图在线预测流程,构建基于CNN-LSTM的驾驶意图单步预测深度学习模型,提出驾驶意图多步递归预测方法;其次,考虑到发动机最大效率区与尾气最小排放区不一致,建立ICHEV等效尾气排放模型,提出基于EEMS的能量在线管理策略;再次,分析驾驶意图频繁变化对EEMS的扰动影响,设计基于MPDI的整车需求功率求解算法,给出EEMS策略的哈密顿函数求解方法。实验结果表明,基于MPDI的EEMS策略可有效降低ICHEV尾气排放量,与传统能量管理策略相比平均降低了25%。Vehicle"carbon compliance"and"carbon neutralization"are important components to achieve the carbon compliance of the whole society.To reduce the total emissions of Intelligent Connected HEV(ICHEV),an Equivalent Emissions Minimization Strategy(EEMS)of ICHEV based on multi-step prediction of driving intention(MPDI)is proposed.First,the online prediction process of driving intention under V2X is presented,a deep learning model of CNN-LSTM is constructed to make single-step prediction of driving intention,and a multi-step recursive prediction method of driving intention is proposed.Second,considering the inconsistency between the maximum efficiency zone and the minimum tailpipe emission zone of engine,an equivalent emission model of ICHEV energy consumption is established,and an energy online management strategy based on EEMS is proposed.Third,the disturbance effect of frequent change of driving intention on EEMS is analyzed,the algorithm for solving the whole vehicle demand power based on multi-step prediction of driving intention is given,and the Hamiltonian function solution method of EEMS is given.The experimental results show that the EEMS based on multi-step prediction of driving intention can effectively reduce ICHEV exhaust emissions by 25%on average compared with the traditional energy management strategy.

关 键 词：智能网联HEV 能量管理策略 驾驶意图预测 等效尾气排放 

分 类 号：TN915[电子电信—通信与信息系统]