Modeling And Energy Management Strategy For Vehicular Hybrid Energy Storage System

Plug-in Hybrid Electric Vehicle(PHEV)is an effective way to resolve the currently fossil energy over-sonsuming and air pollution.In order to ensure the efficient application and long lifespan of PHEV in random and uncertain environment,the following researchs for energy management of Hybrid Energy Storage System(HESS)are carried out in this paper.(1)The DC/DC converter efficiency experiments under different temperatures,switching frequencies,duty cycle and switching device materials,the battery and ultracapacitor characteristic experiments under different temperatures,and different types of ultracapacitor accelerated life experiment under different temperatures and cut-off voltages are correspondingly performed.An abundant experimental dataset is accumulated,and the characterizations of the electrical-thermal-energy for the key components of HESS are analyzed in details,and the key point of the HESS modeling is focused.(2)Because the efficiency of DC/DC converters in HESS is difficult to calculate accurately under multiple factors,an efficiency modeling method for DC/DC converter considering multiple factors is proposed.The track of DC/DC converter under different factors is researched based on the experimental dataset.After the full analysis of the conduction loss and switching loss,the efficiency modeling method with the MOSFET Si C or IGBT Si switching devices under different factors is studied,respectively.The results show that the efficiency model has good accuracy under different factors,which will provide solid data support for subsequent research on energy management strategy.(3)The different ambient temperatures can significantly impact the accuracy and adaptability of ultracapacitor equivalent circuit model,and a data-driven ultracapacitor model parameter identification method based on genetic algorithm with the evaluation indicators of model accuracy and adaptability is proposed.A systematic analysis of precise ultracapacitor modeling method considering ambient temperatures is realized.According to the establishment of five typical ultracapacitor equivalent circuit models,the model accuracy,adaptability and performance under different SOC ranges are fully analyzed.After the verification of ultracapacitor bank model,the ultracapacitor SOC ranges of [0.5,1]is the high efficiency application area.In addition,the Thevenin model is more suitable for ultracapacitor modeling on the basis of the comprehensive consideration on maximum errors,mean errors,RMS errors model complexity and adaptability.(4)The peak components and transient changes in power demand will seriously affect the lifetime of HESS.Firstly,the basic principle of wavelet transform algorithm is clarified.The model of HESS is established based on a data-driven method based on the three dimensional response surfaces among battery or ultracapacitor model parameters,ambient temperatures and SOC.After the analysis among highway road(HWFET),suburb road(WVUBUS)and urban road(MANHATTAN),the wavelet transform with 3 decomposition levels is preferred for the ernegy management via the evaluation indicators of energy loss and adaptability.Moreover,three off-line energy management strategies based on wavelet transform,wavelet transform and rules,wavelet transform and fuzzy considering the ambient temperature factors are proposed.After the analysis of sensitivity of energy management strategy to the number of the hidden layer nodes and the width of the rolling window,a wavelet neural network short term prediction energy management strategy considering the ultracapacitor durability is further presented.The simulation results show that the wavelet transform,which can effectively avoid the impact of peak components and transient changes in power demand on the HESS,has a good effect in off-line and prediction energy management strategy of HESS.(5)In the view of the application and verification for the wavelet energy management strategy under multiple factors,the x PC target based HESS hardware-in-the-loop(HIL)test bench is established.The performances of different wavelet energy management strategies are carried out via HIL test platform for evaluation and verification.The results show that the wavelet energy management strategy presented in this paper can avoid the peak components or transient changes in power demand and show good robustness.