Research On Energy Optimization Of Plug-in Hybrid Electric Vehicles Under Connected Environment

With the rapid development of information network technology,the era of Internet of everything is coming,and the development of information technology such as 5G also provides technical guarantee for the transformation and upgrading of traditional vehicles to intelligent networked vehicles.With the continuous breakthrough of information exchange technology between vehicles and the outside world,it also allows vehicles to start to have autonomous optimal decision-making capability.At the same time,intelligent networked hybrid vehicles can effectively reduce energy consumption and environmental pollution,to achieve the carbon peak,carbon neutral "double carbon strategy" is of great significance.To this end,there is an urgent need to develop a set of real-time energy control strategies that meet the collaboration of vehicle,road,cloud,and network to ensure that the vehicle achieves optimal economic driving.In this paper,based on vehicle to everything(V2X)technology,we study the optimal energy control strategy for plug-in hybrid electric vehicles in dynamic traffic in a networked-following scenario.The dynamic traffic in this paper is set as follows: the front vehicle in the netlinked scenario may be affected by traffic jams caused by other vehicles,etc.,and the Signal Phase and Timing(SPa T)information of the signal lights is not known for the whole journey,for the target vehicle with Vehicle to Infrastructure(V2I)information technology For the target vehicle with Vehicle to Infrastructure(V2I),its information transfer can only receive information from the two nearby signals.The details of the study are as follows:First,the power system model of the target vehicle is analyzed,and the power split planetary row structure is used as the power coupling device,and the differences between it and the conventional series and parallel configurations are compared and analyzed.The rotational inertia factor is considered in the planetary row transmission structure modeling,and the model equation is established using backward modeling simulation.And the model of the net-linked road condition and traffic signal is established.Secondly,the speed planning is also a very important aspect for the net-linked vehicles.For the speed prediction of the front vehicle in the net-linked following scenario,radial-based neural network speed prediction,Gaussian process speed prediction and acceleration sequence-based speed prediction methods are proposed and comparative analysis is performed.The comparison results show that the acceleration sequence-based speed prediction has higher prediction accuracy and more accurately predicts the time node when the front vehicle arrives at the approaching signal,which provides information support for speed optimization of the target vehicle in the netlinked following scenario.Again,the Pontryagin’s Minimum Principle(PMP)based on Model Predictive Control(MPC)is used to solve the problem of real-time solution for the energy management policy solution of the net-linked vehicles,and is combined with Dynamic Programming(DP)methods are studied in comparison.The existing reference battery State of Charge(SOC)curves often adopt the strategy of equidistance assignment,which does not consider the optimal variation of SOC under different slopes.Compared with the equidistant allocation strategy,the results fit the SOC curves based on DP more closely and have better fuel economy.Finally,in the net-linked following scenario,multiple objectives of following passage efficiency,following safety and fuel economy are considered to plan the speed profile of the target vehicle,and the effects of different optimal constant speed settings in the speed planning algorithm on the total fuel consumption are analyzed.The PMP+MPC method with Bayesian optimization is used to solve the energy management strategy,and simulation analysis is performed.