Research On Energy Management Strategy Of 4WD Hybrid Electric Vehicle Based On MPC

Control strategy is one of the key technologies in the field of hybrid electric vehicle.Because of its important influence on the fuel economy,power and emission performance,control strategy has become a research focus in recent years.Especially,improving fuel economy is the moving direction of control strategy of hybrid electric vehicles.In this paper,a four-wheel drive hybrid electric vehicle(4WD HEV)is taken as the research object and fuel economy is considered,and the research mainly focuses on the application of model predictive control in this 4WD HEV.The forward simulation model of the 4WD HEV was set up based on the establishment of the key components model and the vehicle longitudinal dynamics model.The rulebased control strategy was developed to verify the validity of the simulation model.The global optimization strategy based on dynamic programming(DP)was built to get the global optimum and ensure the efficient work of engine in the meantime.The results of these two control strategies provide theoretical basis and performance benchmark for the design of subsequent model predictive control strategy(MPC).By combining DP with rolling optimization,the MPC strategy was designed and several key problems in the process of optimization were studied emphatically.In order to slow down the decline rate of SOC,the penalty function was introduced and the strategy was improved.The fuel consumption of MPC strategy under different prediction horizon was studied to verify the effectiveness of rolling optimization method.Then,the multi-step Markov model was selected to predict velocities and accelerations by analyzing the prediction effects of three kinds of prediction models and comparing the root mean square error of prediction results,and a stochastic model predictive control(SMPC)strategy is formulated.The operating area of engine was optimized based on the simulation results of SMPC to improve the working efficiency of engine.Finally,through comparing the simulation results of SMPC,the frozen-time MPC(FTMPC)and the prescient MPC(PMPC),the stochastic predictive method was proved feasible and the SMPC strategy was certified better in fuel economy and control performance than the ruled-based strategy.