Improved Backward Smoothing—Square Root Cubature Kalman Filtering and Variable Forgetting Factor—Recursive Least Square Modeling Methods for the High-Precision State of Charge Estimation of Lithium-Ion Batteries

Abstract

Accurate lithium-ion battery charge state estimation is crucial for battery management systems. For the modeling of dual polarization - electrical equivalent circuit based on ternary lithium batteries, a variable forgetting factor recursive least square method is proposed for parameter identification given the insufficient tracking ability of the traditional recursive least squares method for abrupt and time-varying signals in a non-stationary environment. A backward smoothing square root cubature Kalman filtering algorithm is applied to enhance the accuracy and convergence speed of SOC estimation. The algorithm uses the square root update to ensure the numerical stability of the filtering and uses the idea of backward smoothing-forward filtering to improve the filtering accuracy on the basis of the first forward filtering. Finally, variable forgetting factor recursive least square is combined with backward smoothing square root cubature Kalman filtering to achieve the joint estimation of model parameters and state of charge, and the feasibility of the battery state of charge estimation is verified in different working conditions. The simulation results show that the variable forgetting factor recursive least square-backward smoothing square root cubature Kalman filter algorithm improves the study's filtering accuracy and convergence speed of lithium-ion batteries.

Modeling an Adaptive Hybrid Soft Sensor with Co-training Learning toward Applications in Wastewater Treatment

“…To adapt to new conditions, the popular solution is to use the adaptive method, such as recursive models, time differential models, moving window methods, and just-in-time learning . The most widely used method is the recursive model , because this can not only allow the model to capture the information on recent data but also renew the original recursive model with less computational intensity.…”

State of Charge Estimation Method of Energy Storage Battery Based on Multiple Incremental Features

“…Increasingly, researchers utilize Kalman filters and their variants to enhance the accuracy of SOC estimation. [12][13][14][15][16] While the model-based approach combined with Kalman filtering has achieved a certain level of estimation accuracy, its performance heavily relies on the accuracy of the model and its parameters. These parameters exhibit significant variations under different operating conditions, necessitating a substantial amount of experimental data to establish precise model parameters.…”

A Prediction Framework for State of Health of Lithium-Ion Batteries Based on Improved Support Vector Regression

“…More complicated models, closely related to physics and electrochemistry principles, have also been used for SOH prediction. The modelbased optimal state prediction methods, such as least square, 12 Kalman filter 13 and particle filter 14 are also adopted to predict battery parameters by minimizing the model output error. Hu et al 15 used an adaptive infinite Kalman filter to predict the state of charge (SOC) of lithium-ion battery, and systematically evaluated its performance under large initial error, wide temperature range and different driving cycles.…”