Research On Multi-objective Energy Management For Connected And Automated Hybrid Electric Vehicles

In the context of the energy crisis,environmental pollution,and stringent restrictions on vehicle exhaust emissions,hybrid electric vehicles exhibit significant advantages in improving fuel economy and reducing emissions.In the meantime,the rapid development of intelligent transportation system has fostered the network connection of hybrid electric vehicles.Therefore,the overall performance of hybrid electric vehicles can be greatly improved by combining vehicle networking technology with energy management technology.In this paper,the multi-objective energy management optimization considering pollutant emission is carried out for a connected hybrid electric vehicle in car-following scenarios,and the specific research contents are as follows:(1)Energy Management Strategy(EMS)of a parallel hybrid electric vehicle based on the Dynamic Programming(DP)algorithm is studied.Based on the ADVISOR database,the mathematical models of the main power components and the vehicle longitudinal dynamics are established.The basic principles of DP algorithm are elucidated,and the cost function aiming at improving fuel economy is defined so as to construct the DPbased EMS.Then,the global optimum is obtained through optimization based on the established model and compared with the results using a traditional controller in terms of the fuel economy.(2)A comparative study on multi-objective EMS of connected hybrid electric vehicles based on Model Predictive Control(MPC)in car-following scenarios is carried out.The theory of MPC is described at first and then,based on such a theory,a multi-objective EMS optimization framework is formulated,which considers inter-vehicle safety,riding comfort,and fuel economy.To find the most fuel-efficient method,two approaches,including the traditional Adaptive Cruise Control(ACC)in series with EMS(ACC +EMS)method,ACC and EMS fusion(ACC-EMS)method,are used to obtain the optimal solution.A comparative analysis of these two strategies reveals the advantages of the ACC-EMS strategy in improving fuel economy.(3)A comparative study of the multi-objective optimization considering exhaust emissions and energy consumption based on MPC is carried out.The multi-objective optimization framework of ACC-EMS,which considers the engine emission model,is constructed based on MPC.The emission weight coefficient corresponding to the optimal tradeoff between energy consumption and emission is determined to improve the comprehensive vehicular performance in car-following scenarios.In addition,a multiobjective EMS based on Predictive Cruise Control(PCC)(PCC-EMS)is designed in the vehicle following scenario,and the optimal results based on PCC-EMS are compared to that based on ACC-EMS with respect to optimization performance.(4)Research on adaptive adjustment of ECMS optimization strategy based on driving condition identification is conducted.The correlation analysis on the characteristic parameters of the working condition sample database is carried out at first,and then three characteristic parameters are selected for driving condition clustering and driving condition identification.Three categories of typical driving condition data are obtained using the K-means clustering method and such data can be used to acquire offline corresponding equivalent factors and train the condition recognizer.Finally,an adaptive adjustment ECMS energy management strategy is established based on driving condition recognition using LVQ neural network.