State of charge estimation by square root cubature particle filter approach with fractional order model of lithium-ion battery

In this paper, a square root cubature particle filter approach was designed to estimate the state of charge of lithium-ion battery, which not only enhanced the numerical stability and guaranteed positive definiteness of the state covariance, but also increased accuracy and decreased computation quantity. Due to the fractional characteristics of the battery capacitance, a fractional order model was used to formulate the lithium-ion battery. Considering the high accuracy and easy convergence, a particle swarm optimization algorithm was utilized to identify the model parameters. The above-mentioned approach was modelled and translated into C code, which was downloaded into battery control unit of battery management system for experimental validation. Two kinds of dynamic cycles were utilized to validate the proposed approach at different temperatures, where both unscent Kalman filter and cubature Kalman filter were compared with the proposed approach. Experimental results indicate that the proposed approach has better accuracy and robustness, and fractional order model is more accurate than integer order model. Therefore, the square root cubature particle filter with fractional order model of lithium-ion battery is a good candidate to estimate the state of charge.

Automated summary

This paper proposes a square root cubature particle filter approach to estimate the state of charge of lithium-ion battery. By using a fractional order model and particle swarm optimization algorithm, the accuracy and computational efficiency of the estimation are improved. Experimental results demonstrate that the proposed approach has better accuracy and robustness.