Research On Online Energy Management Control Strategy Of PHEV Based On Global Optimization Of Equivalent Factors

As a product of the transition from the hybrid electric vehicle(HEV)to the pure electric vehicle(PEV),plug-in hybrid electric vehicle(PHEV)that can be charged externally to the power grid is assembled with larger capacity battery.Therefore,it combines the advantages of HEV and PEV and it has become a research hot spot in the field of the new energy vehicle.As one of the core technologies of the PHEV,the energy management control strategy directly affects the fuel economy performance of the vehicle.However,the currently-used energy management strategy for PHEV is not robust,and it is difficult to adapt to the requirements of different working conditions and different driving conditions.At the same time,there is still a lot of room for improvement in fuel economy.In this regard,this article combines the national key research and development project undertaken by the research group,and takes a PHEV equipped with a CVT as the research object,and an intelligent control algorithm based on a neural network is introduced into the equivalent fuel consumption minimization strategy to form an online instantaneous optimization energy management control strategy with high robustness and fuel economy is developed.The specific research content includes:1.For a plug-in hybrid electric vehicle equipped with a CVT,the power system configuration is analyzed;a forward simulation model for real-time control is built based on the Matlab/Simulink platform,including PID control principle based driver model,the engine,motor,power battery,CVT numerical model,and vehicle model based on longitudinal dynamics of the vehicle;the working model of the vehicle is analyzed and the mode switching rules are formulated.2.The mathematical model of ECMS is established based on the Pontryagin’s minimum principle;through the numerical fitting of the instantaneous fuel consumption rate of the engine and the instantaneous equivalent fuel consumption rate of the battery,the search space of the optimal control variable is reduced to a limited possible value space,and a real-time approximate equivalent fuel consumption minimum control strategy(A-ECMS)model is established;the key control parameter in A-ECMS equivalent fuel consumption minimization strategy the equivalent factor,is simulated and analyzed,and the results show that the fixed the equivalent factor has insufficient adaptability to PHEV’s different initial conditions and changes in driving conditions.3.To solve the above problems,the dynamic programming(DP)algorithm is introduced to obtain the global optimization sequence of equivalent factors,and an AECMS strategy based on time-varying equivalent factors is proposed.First,based on the DP model,the equivalent factor is used as the control variable,the battery SOC obtained by the A-ECMS algorithm at each moment is used as the state variable,and the engine fuel consumption is minimized as the goal.Then,the DP-AECMS algorithm is designed by embedding the DP into the A-ECMS algorithm.Then,the calculation efficiency,boundary value error and discrete accuracy of the DP-AECMS algorithm are analyzed,results show that the calculation speed and accuracy of the DP-AECMS algorithm have been improved.Finally,DP-AECMS algorithm is used to calculate the equivalent factor sequence of global optimization under NEDC condition,and applied to the simulation of A-ECMS control strategy.The results show that compared to the fixed equivalent factor control strategy,on the DP-AECMS control strategy can effectively improve the fuel economy of the vehicle,which is very close to the globally optimal result obtained by directly applying DP.4.In order to adapt to different driving conditions,an intelligent control method based on BP neural network is introduced to extract the nonlinear relationship between the global optimal equivalent factor and the driving conditions and vehicle state.Then,the DP-AECMS control strategy online application is realized.First,the DP-AECMS algorithm is used to calculate the global optimal sequence of the equivalent factors under three typical working conditions,including congested road conditions,urban road conditions and high-speed road conditions,which are used as the training samples of the neural network.Then,a BP neural network model is built and the sample data is used to train the network,and designed an online controller based on the trained neural network.Finally,the online controller is verified under 10-NEDC operating conditions.The results show that the DP-AECMS online control strategy combined with the neural network can adapt to different driving conditions.Compared with the strategy of directly identifying DP results using the same neural network,the control strategy proposed in this paper has better robustness and the fuel economy has been improved by 2.46%.