Improved Fractional-order AEKF Algorithm SOC Estimation And Hardware-in-the-loop Design

New energy vehicles mainly use lithium-ion batteries as energy sources,and the battery management system is used to manage and monitor the battery packs of new energy vehicles.It can detect various parameters of the battery in real time and control the battery to fully utilize the battery energy while at the same time.,To prevent the battery from being over-discharged when discharging,overcharging when charging,and controlling thermal runaway,ensuring the safe use of the battery,not only can extend the life of the battery,but also improve the efficiency of the battery and extend the car’s life.recharge mileage.The battery SOC value is one of the most important estimation data in the battery management system.Whether it can be accurately estimated directly affects the control and decision-making of the vehicle,and is of great significance to improving the efficiency and safety of the vehicle.This paper proposes an adaptive fractional EKFSOC estimation algorithm with temperature correction,and carries out hardware-in-the-loop design and verification.First of all,this article summarizes the development status and research results of battery modeling and SOC estimation at home and abroad for battery model and SOC estimation.Starting from three aspects of battery model establishment,parameter identification and SOC estimation algorithm,the battery SOC estimation Conduct research and propose improvements.Secondly,a characteristic test experiment was carried out on the ternary lithium-ion battery,considering the influence of temperature changes on the parameters of the established battery model,through continuous and pulse charge and discharge experiments,the relationship between battery characteristic parameters and temperature was established;the battery was subjected to AC impedance Spectral experiment,replace the capacitor in the second-order RC equivalent circuit model with CPE components,establish the RQ fractional equivalent circuit model,use temperature to correct the model parameters,combine the fractional-order theory to discretize the model,and adopt improved genetics The algorithm performs offline identification of the model,and compares the model with the traditional integer-order model under the impulse condition;compares the model with temperature correction with the constant parameter model under the DST condition.The results show that the fractional-order model has better accuracy than the traditional model,and the model with temperature correction has better accuracy than the constant parameter model.Thirdly,the algorithm formula of Fractional Extended Kalman with temperature correction is deduced and improved.The adaptive rules are used to adjust the noise covariance in the extended Kalman.Through simulation,the accuracy and robustness of the adaptive fractional extended Kalman algorithm with temperature correction and other Kalman filter algorithms are compared;through different settings The convergence of the adaptive fractional extended Kalman filter algorithm with temperature correction is compared under the initial value;the accuracy and robustness of the adaptive fractional extended Kalman filter algorithm with temperature correction are compared at different temperatures.Finally,a hardware-in-the-loop system for lithium-ion battery SOC estimation is established,and the real-time collected current,voltage and temperature signals are transmitted to the upper computer,and the T-AFEKF algorithm is embedded in the upper computer to realize the real-time estimation and display of the battery SOC.