Lithium-ion Battery Pack States Estimation And Satefy Management Based On System Complexity Analysis

Nowadays,energy crisis and environmental polllution are two major problems facing by the society.Due to the features of high energy density,long cycle life,low self-discharge rate and non-polluting,the lithium-ion battery has been widely applied in the field of new energy to reduce the energy consumption and environmental pollution.The traditional non-linear analyzation methods cannot meet the application requirement in complex system analyzation caused by the complexity of lithium-ion battery characteristics.Therefore,the methods of power lithium-ion batteries complexity analyzation and modeling,the battery states estimation and safety management are researched to build up a power battery pack system analyzation theory and provide reference for safe and efficient system management based on complexity analysis methods.The main contents and contributions of this thesis can be listed as follows:(1)The methods of power battery system describing and modeling are studied based on complexity analysis theory.The complexity of the battery system is analyzed in terms of the battery reaction mechanism,process characteristics and attributive characteristics.Meanwhile,a complex system analyzation theory in complex system simplifying,model reliability and validity analyzation is set up,which is the foundation of power battery system modeling.(2)The methods of complex system analysis are applied in power battery system modeling by analyzing the influences of battery inconsistency.And on this basis a more accurate battery pack model is built up to reflect the battery pack process characteristics.Furthermore,the approach of EKF-UKF is proposed in online identifying the battery model parameters and predicting the battery pack SOC simultaneously to avoid the effects of measurement and observation noise.(3)A more accurate battery model is built up by analyzing the impacts of different working temperatures and discharge/charge current rates,which can reflect the characteristics of different frequency of the battery pack.A novel definition of battery remaining discharge energy is proposed to reflect the real-time battery characteristics and the function of battery state and observation equations are set up.Besides the method of RLS is applied in battery model parameters identification,while the UKF is used in battery remaining discharge energy estimation.The validity and robustness of the proposed method are verified under the profiles of NEDC and real working conditions.(4)The battery pack SOH can reflect the service condition of battery usage,which affects the battery maintenance cost.In this thesis,a novel definition of battery pack SOE and SOH is proposed based on the change of battery pack maximum energy storage,and a more accurate battery pack model is built up based on the analyzation the change of battery attributive characteristics.To improve the execution efficiency of the algorithm,a novel battery states estimation method is proposed based on multi-time scale in identifying the battery model parameters and predicting the battery pack states.The validity and robustness of the proposed battery model and the veracity of battery SOE and SOH estimation method are verified under different profiles.(5)The methods of battery system identification and states estimation are applied in making up a battery pack and optimizing the discharge/charge strategy,equalization management strategy and battery fault management strategy to improve the safety of battery system,which can be employed in real application.