| At present,plug-in hybrid electric vehicle(PHEV)has gradually become a research hotspot of new energy vehicles.As it has two power sources,the development of energy management strategy plays an important role in excavating the energy saving potential of PHEV and is one of the core technologies of PHEV.In this paper,PHEV is taken as the research object,the goals are to excavate the potential of energy saving and improve the fuel economy of the whole vehicle.Based on the driving condition data of a specific driver,Matlab/Simulink is used as the simulation platform to analyze the driving condition characteristics of the historical data,the reference trajectory planning algorithm of battery state of charge(SOC)based on the global optimization strategy and the adaptive control strategy based on dirving condition prediction research.Firstly,this paper analyzes the different configuration and working mode of PHEV power system.After selecting the configuration,the whole vehicle simulation model is built based on the idea of forward simulation modeling in MATLAB / Simulink,including models of driver,engine,motor,power battery and vehicle dynamic.Then,a rule-based energy management strategy is established,and the effectiveness of the simulation model and the rationality of the control strategy were verified by NEDC driving cycles.Secondly,a specific driver is taken as the tested object to analyze and construct the working condition characteristics based on historical data.The actual road trip data of the driver is collected by the autonomous driving method,and the data collection and preprocessing scheme based on a driver is established.Then the historical condition data is divided into several short condition segments,the characteristic parameters of each condition segment are calculated,and the dimension of the condition characteristic parameters is reduced based on the principal component analysis method.Considering that the fuzzy c-means clustering method is very sensitive to the initial clustering center,a fuzzy c-means clustering method based on genetic simulated annealing algorithm is proposed,So the historical condition of data is divided into four categories,and constructs the driving condition,this all provides the data support to the research of the follow-up control strategy.Thirdly,the global optimal control strategy based on dynamic programming(DP)is established,and the planning method of SOC reference trajectory is proposed.The DP theory is applied to the problem of PHEV energy management,so the control strategy of PHEV based on the DP algorithm is established.Then,taking euclidean distance as the clustering index,the method of driving condition identification based on clustering analysis is proposed.On this basis,the optimal control rules of SOC under different types of typical driving conditions is studied by DP,the reasonable power distribution range of each type of typical working conditions is determined,and the planning algorithm of SOC reference trajectory is proposed.Finally,the adaptive control strategy of PHEV based on condition prediction is studied.The energy management strategy of equivalent consumption minimization strategy(ECMS)is established and the calculation method of adaptive equivalent factor is presented.Then,the research on the method of generating the target driving condition considering the traffic information is carried out.By obtaining the average traffic flow information of the road ahead,the target driving condition that the vehicle may run in the future is generated.At the same time,a speed prediction model based on NAR neural network is proposed to further optimize the constant part of the equivalent factor.Based on this,the adaptive ECMS(A-ECMS)method based on the condition prediction is formed.Through the simulation test of various control strategies,the results show that the oil saving of A-ECMS based on the driving condition prediction is about 8.7% higher than that of the rule-based strategy,so as to improve the fuel economy of the vehicle. |
PDF Full Text Request