The Research On Battery SOC Estimation Algorithm Based On PI State Observer

There are three main technologies in electric vehicles: batteries,motors,and electronic controllers.As the primary or only power-driven energy for electric vehicles,power battery pack is the main bottleneck technology that restricts its development.The Battery Management System(BMS)monitors,evaluates,and diagnoses battery status at all times,which is directly related to the accuracy and safety of battery during operation.And accurate and fast state-of-charge(SOC)estimation is a core technology that is crucial for battery systems and energy management systems.In this paper,the lithium-ion battery is taken as the research object,and the SOC state estimation of the battery is deeply studied.First,the basic working principle of the battery is briefly introduced and its characteristics are analyzed.At present,lithium-ion power batteries have many advantages such as: small self-discharge,good cycle characteristics,and high energy efficiency.Secondly,a variety of equivalent circuit models are established for lithium-ion batteries.The online and offline identification of the battery model parameters are realized through experimental data of Hybrid Pulse Power Characterization(HPPC)Test and Dynamic Stress Test(DST).And the accuracy of each model and the accuracy of online and offline parameter identification is verified.Then,a method for estimating battery impedance and SOC based on multi-level Proportional-Integral(PI)Observer is proposed.The PI observer is capable of suppressing modeling errors and achieving accurate estimation with less computational effort.In addition,the compensation factor ? introduced in the system model can compensate part of the battery attenuation capacity and change the characteristics of the battery model according to the battery usage(characterized by the mean value of the battery impedance)to improve the accuracy of the SOC estimation under multi-state conditions.The effectiveness and applicability of the algorithm are verified by the composite dynamic stress test(DST)experiment.The results show that the proposed method can achieve battery SOC estimation under different degrees of use,and suppress errors caused by noise such as measurement.During the effective discharge period,the maximum estimation error of the SOC can be kept at around 2%.Finally,the simulation of the PI state observer based on Thevenin/Two-Order RC model is carried out,and the feasibility of the model compensation factor ? under different system models is analyzed.The results show that the introduction of the compensation factor ? can adapt to the SOC estimation under different degrees of use of the battery,and also enable the system to tolerate a certain degree of impedance fluctuation and capacity attenuation under the premise of accurate SOC estimation.The fault-tolerant area of the compensation factor will gradually increase as the battery usage becomes deeper,and will become narrower due to the improved accuracy of the system model.