Power Allocation Smoothing Strategy For Hybrid Electric Vehicle Based On Markov Algorithm

Electric vehicles have the advantages of zero emissions,low power consumption,low noise and easy to maintain.In the current environment of increasingly serious environmental pollution and the gradual depletion of fossil energy,it is necessary to replace petrol vehicles with electric vehicles.However,there are still many obstacles to overcome in the large-scale commercial application of electric vehicles,such as shorter cruising range,higher energy storage system cost,longer charging time and shorter battery cycle life.To solve the above problems fundamentally,it is necessary to find a breakthrough from the energy storage device of electric vehicle,and the hybrid energy storage system composed of a lithium battery and an ultracapacitor is applied to the electric vehicle,which provides a possibility to solve the above problem.Therefore,this paper takes energy management of hybrid energy storage system in electric vehicle as research content,considers energy management strategy optimization and ultracapacitor sizing optimization as a whole to realize the integrated optimization of hybrid energy storage system.Firstly,models for electric vehicles and hybrid energy storage systems were established.Lithium battery/ultracapacitor model and DC/DC converter efficiency model are used to calculate the energy loss in lithium battery,ultracapacitor and DC/DC converters.An electric vehicle dynamics model for calculating power demand based on speed data of driving cycle.Finally,a dynamic battery capacity degradation model was established to analyze the impact of energy management strategy on battery cycle life.Secondly,this paper uses bilinear interpolation to smooth the energy management strategy based on Markov decision process.In this way,the power fluctuation can be mitigated meanwhile the computation cost is not greatly increased.Considering the energy loss and the energy reserve in a HESS,the reward function is built.Considering the different operating characteristics of the lithium battery and the supercapacitor,two types of DC/DC converters are used to make full use of the capacity of the supercapacitor while minimizing the hardware cost of the DC/DC converter.Utilizing the cumulative reward function,the effect of UC pack sizing on the performance of power allocation is analyzed in detail,the appropriate UC sizing can be obtained further.Finally,Simulation results show that MDP strategy with the bilinear interpolation can not only reduce the energy loss by 5~10%,but also prolong the battery cycle life.In addition,this paper also builds a scaled-down platform of the hybrid energy storage systems,using the master-slave control for controlling the hybrid energy storage system.The experimental results verify the effectiveness of the proposed energy management method.