Research On Energy Management Strategy Of Plug-in Hybrid Electric Vehicle In Car-following Environment

With the environmental pollution and energy shortage becoming more and more serious,the automobile industry is moving towards the goal of energy saving and environmental protection.Developing new energy vehicles is the key way to achieve the above goal.Plug-in Hybrid Electric Vehicle(PHEV)has become the object of vigorous development in our country because it can be charged by external power supply and has a large battery capacity,so its endurance mileage has increased a lot compared with the traditional hybrid electric vehicle,and there is no concern about the power consumption of electric vehicles.On the other hand,with the rapid development of intelligent interconnection technology,more and more automobiles are equipped with Internet terminal technology.In more and more complex traffic environment,mutual assistance and cooperation among vehicles to realize the driverless function of some vehicles is the only way to realize self-driving of a single vehicle.In addition,the energy management strategy is the core technology of PHEV,which directly affects the performance of PHEV.In this paper,the energy management strategy of PHEV in automobile environment is studied to make it have better safety,comfort and fuel economy.The main research contents are as follows:(1)The structure of parallel PHEV power system is analyzed,and the power transmission system model is established.Based on the experimental data,the efficiency model of engine,motor and power battery is established.In the car-following environment,the Grey Neural Network(GNN)is used to predict the future speed trend of the front vehicle.The influence of the historical speed time window and the length of prediction domain on the prediction accuracy of GNN is analyzed,and the optimal historical speed time window and prediction domain are determined.On the basis of the future speed trend of the front vehicle,a variable time-distance car-following control strategy considering the future speed trend of the front car is proposed.Firstly,with the goal of safety and comfort,the expected time-distance is solved by using the variable time-distance control algorithm,and then the expected speed of the PHEV is programmed by using the fuzzy adaptive PID control algorithm to achieve multi-objective adaptive car-following control.(2)For PHEV,the SOC reference trajectory is needed when formulating energy management strategy.Firstly,typical working conditions are extracted based on K-means clustering for mixed working conditions including Chongqing working conditions.Then,based on dynamic programming algorithm(DP),the influencing factors of SOC decline rule are analyzed,and the characteristic parameters representing SOC decline rule are extracted offline and applied to typical working conditions.In online application,the current working conditions are identified based on the results of K-means clustering,and then the off-line calculated reference SOC variation corresponding to typical working conditions is selected as the current SOC reference.In addition,the probable variable travel problems of vehicles is discussed,and a method of planning SOC reference variation when the driving mileage changes is proposed.The simulation results show that the correlation coefficient between the reference trajectory and the theoretical SOC trajectory calculated by DP algorithm reaches 99% under the premise of the significance test p value is zero,showing a strong correlation characteristic.(3)On the basis of the reference SOC variation,an Equivalent Fuel Consumption Minimum Strategy(ECMS)based on fuzzy control to adjust the equivalent factor is proposed to solve the problems of difficult determination of the optimal equivalent factor and poor adaptability to operating conditions in the ECMS.Firstly,the influencing factors of equivalent factor are analyzed,and the optimal equivalent factors under 25 combined conditions are calculated iteratively by genetic algorithm,and the initial equivalent factor query map is formulated.In addition,the possible range of equivalent factors is analyzed,which shortens the searching time of genetic algorithm.Finally,the equivalent factors are adjusted in real time by using fuzzy control.The simulation results of two cycles show that the energy management strategy achieves 98.04% and 97.81% of the theoretical results of DP respectively,and the total cost increases by 1.37% and 1.27% compared with the PI-adjusted equivalent factor ECMS strategy.(4)The hardware-in-the-loop(HIL)based on D2P-NI PXI is carried out for the proposed energy management control strategy in car-following environment.The model of vehicle and controller is established,the overall scheme of experiment is determined,the experiment platform is integrated,and the experiment simulation platform is built.The experimental results show that the total cost of energy management strategy proposed reaches 95.5% of the theoretical result of DP,and good safety and comfort are guaranteed in the process of car following.