Research On Multi-objective Control Strategy Of Plug-in Hybrid Logistics Vehicle Based On Driving Cycle Prediction

With the rapid development of the logistics industry,logistics vehicles have brought about energy consumption and environmental pollution hazards while contributing to urban economic development.Plug-in Hybrid Logistics Vehicle(PHLV)has received widespread attention because of its high level of energy saving and emission reduction,large battery capacity and external charging,which is suitable for inter-city logistics transportation with long transport distance.Energy Management Strategy(EMS),as the core technology of PHLV,plays a decisive role in the performance of the vehicle,and the existing research mainly focuses on EMS that only considers fuel economy.With the increasing stringency of emission regulations,it is becoming more and more difficult to meet the requirements of EMS that only considers fuel economy.In addition,existing research shows that future driving cycles will also have an impact on the performance of the vehicle.Therefore,it is of great practical significance to study the multi-objective control strategy of PHLV considering fuel economy and emission considering the influence of future driving cycles.This paper focuses on the scientific problem of"multi-objective control strategy of plug-in hybrid logistics vehicle based on driving cycle prediction",and the following work has been accomplished:(1)PHLV power system modeling and rule-based EMS research.The configuration characteristics of PHLV,the subject of this paper,are analyzed and the operating mode is classified;the modeling idea of forward simulation is used to establish the whole vehicle model of PHLV on Matlab/Simulink platform,including vehicle longitudinal dynamics model,driver model,and powertrain component model;the rule-based CD-CS-ICE(Charge Depleting-Charge Sustaining-Internal Combustion Engine)energy management strategy was designed and implemented on Matlab/Simulink platform;the fuel economy and emission performance of the PHLV under the CD-CS-ICE energy management strategy were simulated under 10 CHTC-LT cycles,which verified that the PHLV vehicle model and CD-CS-ICE energy management strategy established in this paper are reasonable and effective,and laid the foundation for the optimization design of the energy management strategy in the later paper.(2)Research on the future short-term driving cycle prediction method of PHLV.By analyzing the characteristics and application scenarios of the current common driving cycle prediction methods,the driving cycle prediction models based on BP and RBF neural networks are selected for the short-term driving cycle prediction research in this paper;according to the characteristics of PHLV,the corresponding training and validation sample data are selected and the evaluation index of the driving cycle prediction models are given;after determining the topology and a series of relevant parameters of the BP and RBF neural network driving cycle prediction models,a total of 27 driving cycle prediction models with different prediction horizons length were trained and compared,and finally two models with excellent prediction performance but different prediction horizons length,BP-5-5-15 and BP-5-10-20,were selected for the subsequent vehicle speed prediction;in order to make comprehensive use of BP-5-5-15 and BP-5-10-20 prediction models,a variable-horizons driving cycle prediction model was established and validated by making the switching rules of prediction models,the root mean square error RMSE of the variable-horizons driving cycle prediction model was reduced by 13.61%and 7.68%compared with the fixed-horizons BP-5-10-20 prediction model under the validation sample cycle and CHTC-LT cycle.The establishment of variable-horizons driving cycle prediction model also lays the foundation for the development of PHLV multi-objective control strategy incorporating short-term driving cycle prediction in the later paper.(3)Research on the multi-objective control strategy of PHLV based on driving cycle prediction.In order to improve the energy saving and emission reduction level of PHLV,based on the CD-CS-ICE energy management strategy,a multi-objective optimization problem is constructed by establishing a cost function with equivalent fuel consumption as fuel economy index and CO,HC and NO_x emissions as emission index in the CS and ICE stages after the battery SOC is reduced to 0.3 for the first time;an adaptive adjustment method of the equivalence factor based on the combination of variable-horizons driving cycle prediction model and battery SOC feedback is designed;finally,the multi-objective optimal control problem of PHLV is solved and verified by simulation based on the principle of Adaptive Equivalent Consumption Minimization Strategy(A-ECMS),compared with the basic CD-CS-ICE strategy,the designed variable-horizons multi-objective control strategy(A-ECMS with Variable horizons,VAECMS strategy)achieves3.15%reduction in fuel consumption per 100 km,5.35%reduction in CO emission per km,6.99%reduction in HC emission per km,11.48%reduction in NO_x emission per km,and the fixed-horizons multi-objective control strategy(A-ECMS with Fixed horizons,FAECMS strategy)achieves 2.14%reduction in fuel consumption per 100 km,0.31%reduction in CO emission per km,3.16%reduction in HC emission per km,1.22%reduction in NO_x emission per km,the improvement effect of VAECMS strategy was better than that of FAECMS strategy.