The Research On Model Predictive Control Strategy Of Plug-in Hybrid Electric Vehicle

Since 21 century,energy shortages and environmental protection have become the hot topics in the whole world.With the rapid development of the automobile industry,automotive manufacturers of the major countries around the world have realized the importance of the development in the new energy vehicle industry.Among the various types of new energy vehicles,plug-in hybrid electric vehicle with the advantages of both pure electric and conventional vehicle is a more practicable selection.In this paper,the power system of plug-in hybrid electric vehicles is taken as the research object,and the parameter matching of key components of the power system and the formulation of energy management strategy are the themes of the paper.The aspects of this study are as following:1)By defining the basic structure of the power system with the common structure and vehicle design requirements for plug-in hybrid electric vehicles,the structure of the power system is determined and the parameters of the main power system components such as engine,motor and battery are matched.Then a simulation model of plug-in hybrid electric vehicle was established basing on Simulink/Matlab,which including the key component models of the power systems and the rule-based energy management strategy proved the feasibility of the parameters eventually.2)In order to avoid the imbalance of dynamic performance and fuel economy in the process of parameter matching,the multi-objective genetic algorithm is used to synchronously optimize the transmission system parameters and energy management parameters,ensuring the comprehensive optimization of power performance and fuel economy.3)The study takes the torque distribution of the energy management strategy as an optimization issue in multi-stage decision,the offline optimal torque allocation strategy is solved based on Bellman's principle,then the global energy management strategy based on dynamic programming algorithm is established.4)Taking the dynamic programming as a solver for model prediction rolling optimization,the energy management strategy based on the model prediction is established.And point out the influence of various parameters on fuel economy in the energy management strategy of model prediction.Established an energy management strategy by speed prediction basing on Markov's speed forecast.Adjusting the method of markov prediction to decrease the cumulative error of prediction,so as to improve the fuel economy of the energy management strategy.Finally,The engine workspace of energy management strategy is modified by using the logical threshold so that the fuel economy of the energy management strategy is further optimized.