| As global environmental pollution becomes increasingly severe and traditional fossil fuels gradually deplete,countries around the world are placing high importance on the research and promotion of new clean energy sources.Lithium-ion batteries,as a new type of clean energy,have the advantages of high energy density,light weight,long service life,and no memory effect.Therefore,they are widely used in portable electronic devices,electric vehicles,and energy storage systems.The State of Charge(SOC)of lithium-ion batteries is an important parameter during the working process,and obtaining high-precision SOC estimation results in practical working environments is of great significance for improving battery performance,extending life,and enhancing safety.Traditional SOC estimation algorithms are mostly applicable for estimating SOC under Gaussian noise,but in actual working environments,lithium-ion batteries are often disturbed by non-Gaussian noise.Under such conditions,traditional SOC estimation algorithms are unable to accurately estimate SOC.Therefore,to address the problem of low SOC estimation accuracy under non-Gaussian noise interference,this paper combines the Maximum Correlation-entropy Criterion with the Extended Kalman Filter algorithm to propose two new SOC estimation algorithms.The main research content and innovation points of this paper are summarized as follows:Firstly,the paper describes the structure and working principle of lithium-ion batteries and explains the mapping relationship between lithium-ion impedance characteristics,state of charge(SOC),and open circuit voltage(OCV).The SOC-OCV curve is fitted using a polynomial,and the Thevenin equivalent circuit model and the second-order RC equivalent circuit model are chosen as the research objects for comparison.After identifying the unknown parameters of the models and verifying their accuracy,the state and observation equations of the models are derived to establish the foundation for subsequent SOC estimation.Then,in order to address the issue of low estimation accuracy of traditional Kalman filter algorithms under non-Gaussian noise interference,this paper proposes a Maximum Correlation-entropy Criterion Extended Kalman Filter(MCC-EKF)based on the maximum correlation entropy criterion(MCC).MCC-EKF replaces the Minimum Mean Squared Error(MMSE)criterion of traditional Kalman filter algorithms with MCC,and uses local optimization instead of global optimization to reduce estimation errors caused by nonGaussian noise interference.Experimental results show that the MCC-EKF algorithm has high accuracy in estimating SOC,and exhibits a certain level of robustness,making it an effective method for SOC estimation.Experimental results show that the SOC estimation accuracy of MCC-EKF algorithm is 51.12% higher than that of EKF under the interference of non-Gaussian noise.It can converge to the true SOC value within 10 s if the initial SOC value is set incorrectly.It is an effective SOC estimation method with good robustness.Finally,to further improve the accuracy of SOC estimation under non-Gaussian noise interference,a Maximum Correlation-entropy Criterion Adaptive Extended Kalman Filter(MCC-AEKF)is proposed based on the MCC-EKF algorithm.The MCC-AEKF algorithm updates the process noise variance Q with an adaptive covariance matrix to further reduce the estimation error of the battery state.Experimental results show that,under the influence of non-Gaussian noise,based on No.1 and No.2 battery data,the estimation accuracy of MCC-AEKF algorithm is 24.2% and 22.0% higher than that of MCC-EKF algorithm.MCCAEKF algorithm can accurately estimate SOC at different temperatures.In addition,it can converge to the true SOC value within 30 seconds if the initial SOC value is set incorrectly.Therefore,it has good robustness and is an effective SOC estimation algorithm.This paper proposes two novel algorithms aimed at addressing the issue of inadequate SOC estimation precision for lithium-ion batteries in practical scenarios.These algorithms are capable of providing more precise SOC estimation results,which can facilitate the optimization of battery design and control,improve energy utilization efficiency and cycle life,and enhance the management and maintenance of batteries.This research has significant implications. |
