基于GRU-RBFNN车速预测的A-ECMS能量管理策略  

ENERGY MANAGEMENT STRATEGY OF A-ECMS BASED ON GRU-RBFNN SPEED PREDICTION

作　　者：李昕光 王文超 元佳宇 Li Xinguang;Wang Wenchao;Yuan Jiayu(School of Mechanical and Automotive Engineering,Qingdao University of Technology,Qingdao 266000,Shandong,China)

机构地区：[1]青岛理工大学机械与汽车工程学院,山东青岛266000

出　　处：《计算机应用与软件》2025年第3期34-40,共7页Computer Applications and Software

基　　金：山东省自然科学基金项目(ZR2020MG017)。

摘　　要：为进一步提高混合动力汽车的燃油经济性,提出一种基于车速预测的自适应等效燃油消耗最小策略(Adaptive Equivalent Consumption Minimization Strategy,A-ECMS)。应用VISSIM软件建立实地微观交通仿真模型并获取交通信息,基于PyTorch框架搭建考虑时空特征的门控循环单元-径向基神经网络预测模型。在MATLAB/Simulink/Stateflow中建立混合动力汽车动力学模型,对基于车速预测的A-ECMS与固定等效燃油消耗最小策略(F-ECMS)进行对比研究,仿真结果表明,A-ECMS相较于F-ECMS,SOC波动更小,汽车燃油经济性提升8.97%。In order to improve the fuel economy of hybrid electric vehicles,an adaptive equivalent consumption minimization strategy(A-ECMS)based on vehicle speed prediction is proposed.The VISSIM software was used to establish a microscopic traffic simulation model in the field and obtain traffic information.Based on the PyTorch framework,a gated recurrent unit-radial basis function neural network prediction model considering spatio-temporal characteristics was constructed.The dynamic model of parallel hybrid electric vehicle was built in MATLAB/Simulink/Stateflow,and a comparative study was carried out between the A-ECMS based on vehicle speed prediction and the fixed equivalent consumption minimization strategy(F-ECMS).The results show that compared with the F-ECMS,the SOC fluctuation of A-ECMS is smaller,and the automobile fuel economy is improved by 8.97%.

关 键 词：门控循环单元 径向基神经网络 车速预测 并联式混合动力汽车 等效燃油消耗最小策略 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术] U461.8[机械工程—车辆工程]