摘要
为提升混合动力汽车(PHEV)燃油经济性,提出了一种基于LSTM深度学习的车速预测方法,并在预测车速基础上,以燃油消耗量最少为目标,借助动态规划(DP)算法最优化分配发动机和电机扭矩,从而建立LSTM-DP整车能量管理策略。Matalb/Simulink仿真平台验证结果表明:在实际道路工况下,整车燃油经济性优化了9.37%。
In order to improve the economy of Plug-in Hybrid Electric Vehicle(PHEV),use the Long Short Term Memory(LSTM)neural network to predict vehicle speed.Then,based on the future speed,use the dynamic programming algorithm to assign torque between energy and motor,aim for the optimal fuel consumption.At last,based on the Matlab/Simulink emulation model,in real-time road conditions,the LSTM-DP energy management can induce 9.37%fuel consumption.
作者
马尧
王新树
上官兴兴
边立健
Ma Yao;Wang Xin-shu;Shangguan Xing-xing;Bian Li-jian(Wenzhou Polytechnic,Zhejiang Wenzhou 325006;Chery New Energy Vehicle Co.,Ltd.,Anhui Wuhu 241009)
出处
《内燃机与配件》
2025年第18期12-15,共4页
Internal Combustion Engine & Parts
基金
温州市基础性公益科研项目(G2023078)。
关键词
混合动力汽车
能量管理策略
车速预测
Plug-in hybrid electric vehicle
Energy management strategy
Vehicle speed prediction
