基于车速预测的PHEV预测能量管理策略  

PHEV Predictive Energy Management Strategy Based on Vehicle Speed Prediction

作　　者：魏丽青[1] 强永军 WEI Liqing;QIANG Yongjun(Leshan Vocational and Technical College,Leshan 614000,China)

机构地区：[1]乐山职业技术学院,四川乐山614000

出　　处：《车用发动机》2024年第2期83-92,共10页Vehicle Engine

基　　金：四川省教育厅研究项目(17ZB0199)。

摘　　要：为解决插电式混合动力汽车预测能量管理策略中车速预测不准确导致车辆燃油经济性降低问题,提出了一种基于麻雀搜索算法优化变分模态分解和长短时神经网络的车速组合预测模型。在模型预测控制架构下采用该预测模型对未来车速进行预测,将全局优化问题转换为预测时域内动力源扭矩优化分配问题,以发动机油耗最小为优化目标,采用动态规划算法对预测时域内的优化问题进行求解。通过仿真表明,所提出的组合预测模型较之于LSTM预测模型预测精度提升了59.57%。同时,基于组合预测模型的预测能量管理策略相较于基于LSTM预测模型的预测控制策略燃油消耗降低了4.58%,相较于基于规则的策略燃油消耗降低了15.1%。To address the issue of poor fuel economy induced by the inaccurate vehicle speed prediction in the predictive energy management strategy of plug-in hybrid electric vehicle,a vehicle speed combination prediction model based on optimizing variational mode decomposition and long-term and short-term neural networks with the sparrow-search algorithm was proposed.Under the model predictive control framework,the prediction model was used to predict the future vehicle speed,and the global optimization problem was transformed into the optimal distribution problem of power source torque in the prediction time domain.With the minimum fuel consumption as the optimal goal,the dynamic programming algorithm was used to solve the optimization problem in the prediction time domain.Compared with LSTM prediction model,the proposed combined prediction model improved the prediction accuracy by 59.57%according to the simulation,and the predictive energy management strategy reduced fuel consumption by 4.58%.Compared with the rule-based strategy,it reduced fuel consumption by 15.1%.

关 键 词：插电式混合动力汽车 预测控制 能量管理 车速预测 

分 类 号：U469.72[机械工程—车辆工程]