Research On Estimation Methods For State Of Charge And Health Of Lithium-ion Batteries

In recent years,the rapid development of the new energy industry has alleviated the shortage of traditional fossil fuels and environmental pollution to a certain extent,which helps to achieve sustainable economic development.Lithium-ion batteries are widely used in energy storage power systems and new energy vehicles due to their unique performance advantages,but their safety issues are still the main factor restricting current development.Accurate estimation of State of Charge(SOC)and State of Health(SOH)is of great significance for battery safety management,and has become a current research hotspot,this article focuses on the estimation of SOC and SOH for lithium-ion batteries.This article conducts modeling analysis based on the second-order RC equivalent circuit model of lithium-ion batteries,and compares and analyzes two parameter identification algorithms: Recursive Least Squares(RLS)and Variable Forgetting Factor Recursive Least Squares(VFFRLS).The results show that VFFRLS has higher parameter identification accuracy.Secondly,in terms of SOC estimation,this paper proposes an improved Particle Filter(PF)to address the problems of linearization errors and slow convergence when the initial SOC is incorrect in traditional Extended Kalman Filter(EKF)algorithms.The improved PF introduces the Levy flight strategy into the PF to replace its original resampling process.Through simulation verification under different operating conditions,the results show that the SOC estimation results can quickly converge to the true value of SOC under different initial SOC conditions.For SOH estimation,the EKF algorithm was selected for battery capacity estimation.Simulation results show that accurate capacity estimation results can be obtained even when the initial capacity value is unknown.In addition,for the joint estimation of SOC and SOH,this paper first conducted a joint estimation at a single time scale;Secondly,the number of small fluctuations in battery parameters and capacity over a period of time is classified as parameters at the macro time scale,and SOC is classified as parameters at the micro time scale for joint estimation of SOC and SOH at multiple time scales.The results indicate that joint estimation at single and multiple time scales has lower computational complexity and higher real-time performance when the errors are comparable.Finally,a bidirectional DC-DC microgrid experimental platform containing energy storage lithium-ion batteries was built to stabilize the bus voltage by controlling the charging and discharging of lithium-ion batteries in response to the fluctuation of DC bus voltage.Under this operating condition,perform online parameter identification,SOC and SOH joint estimation for the battery at both single and multiple time scales.The experimental results validate the effectiveness and feasibility of the designed parameter identification and state estimation algorithms.