Research On Energy Management Strategy Of PHEV Based On Model Predictive Control

Plug-in Hybrid Electric Vehicle(PHEV),as a new type of new energy vehicle,has two different power sources,engine and motor.Compared with traditional hybrid vehicles,PHEV is characterized in that its power battery can be charged by an external power source in addition to being charged by the motor during the driving process of the vehicle.Because the battery capacity of PHEV is large,it can realize long-distance pure electric driving.As one of the key technologies of hybrid electric vehicles,energy management strategy undertakes the energy distribution and torque management of the entire system,which is extremely important for the fuel economy and power of the vehicle.This paper takes the plug-in hybrid electric vehicle with single-axle parallel P2 configuration as the research object.Based on a certain PH2 PHEV model,combined with the relevant data obtained from the bench test,a vehicle simulation model was established,an energy management strategy based on certain rules was formulated,and the model was simulated and verified.Aiming at the prediction of vehicle speed,a prediction model based on RBF neural network is established,which can be used in the prediction model of model prediction control in the following.Finally,in the energy management strategy based on model predictive control,combined with dynamic programming algorithm,the energy management problem is optimized.The details are as follows:(1)According to the system structure of plug-in hybrid electric vehicle with P2 configuration,analyze the possible working modes and the corresponding energy flow.According to the knowledge of automobile dynamics and automobile theory,the power requirements of the entire vehicle are calculated,and the parameter matching and selection of each component of the plug-in hybrid electric vehicle are performed according to the performance requirements.(2)Based on the vehicle performance parameters of a certain P2 plug-in hybrid electric vehicle,each key component of the hybrid electric vehicle is modeled to construct a vehicle model framework.On this basis,a rule-based energy management strategy was established,and the power consumption mode and the power maintenance mode were established respectively,and the switching logic and torque distribution of its working mode were modeled.Finally,the rule-based control strategy is simulated and analyzed.The correctness of the vehicle model and the effectiveness of the control strategy are verified.(3)The prediction process of RBF neural network is analyzed,and a vehicle speed prediction model based on RBF neural network is established.Through the training and learning of the vehicle speed training samples,the speed prediction is carried out according to the NEDC cycle conditions.The speed prediction uses different time prediction domains,and finally the prediction results are analyzed.(4)For plug-in hybrid electric vehicles,an energy management strategy based on model predictive control is studied.First,the principle of model predictive control is briefly described,and then a vehicle speed prediction model is established based on RBF neural network.By predicting the vehicle speed in the predicted time domain,the vehicle demand torque in the predicted time domain is obtained indirectly.The dynamic programming method is used to optimize the prediction time domain,and the fuel consumption is used as an index function to complete the reasonable distribution of engine and motor torque.Finally,the strategy is simulated and analyzed to verify the feasibility and effectiveness of the method.