An improved particle swarm optimization-cubature Kalman particle filtering method for state-of-charge estimation of large-scale energy storage lithium-ion batteries

With the global demand for large-scale energy storage strategies, lithium-ion batteries with high energy densities have emerged as the primary energy storage systems. State-of-charge (SOC) is a critical state parameter for energy storage systems that enable safe and effective monitoring of the batte...

Ausführliche Beschreibung

Gespeichert in:

Bibliographische Detailangaben
Veröffentlicht in:	Journal of energy storage Jg. 100; S. 113619
Hauptverfasser:	Wang, Chao, Wang, Shunli, Zhang, Gexiang, Takyi-Aninakwa, Paul, Fernandez, Carlos, Tao, Junjie
Format:	Journal Article
Sprache:	Englisch
Veröffentlicht:	Elsevier Ltd 20.10.2024
Schlagworte:	Cubature Kalman particle filter Forgetting factor-limited memory recursive extended least squares Lithium-ion battery Particle swarm optimization State-of-charge State-of-charge Lithium-ion battery Forgetting factor-limited memory recursive extended least squares Particle swarm optimization Cubature Kalman particle filter
ISSN:	2352-152X
Online-Zugang:	Volltext
Tags:	Tag hinzufügen Keine Tags, Fügen Sie den ersten Tag hinzu!

Beschreibung
Zusammenfassung:	With the global demand for large-scale energy storage strategies, lithium-ion batteries with high energy densities have emerged as the primary energy storage systems. State-of-charge (SOC) is a critical state parameter for energy storage systems that enable safe and effective monitoring of the battery's real-time state. This study proposes an improved particle swarm optimization-cubature Kalman particle filter (PSO-CPF) for SOC estimation of large-scale energy storage lithium-ion batteries. Firstly, this study conceptually combines the forgetting factor and memory length to create the forgetting factor-limited memory recursive extended least square algorithm, which effectively improves the accuracy of online parameter identification and anti-interference. Secondly, for the problems of particle degradation and diversity loss, this study establishes the PSO-CPF model, which effectively improves the particle degradation problem and maintains particle diversity. Finally, to further improve the filtering performance of the model, this study proposes a new fitness function to reduce the impact of noise variance on the final optimized particles. Under complex working conditions of different temperatures, the results show that the maximum error of the improved PSO-CPF is between 1.86 % and 2.84 %, and the mean absolute error and root mean square error are between 0.96 % and 1.19 %, reflecting its good tracking ability. The evaluation metrics show that the proposed model has higher accuracy and better robustness, providing a reference for improving the SOC estimation performance. •Designing the FF-LMRELS algorithm to effectively improve the accuracy of the online parameter identification.•Establishing the PSO-CPF estimation model to effectively improve the particle degradation problem.•Proposing a new fitness function to reduce the impact of noise variance on the final optimized particles.
ISSN:	2352-152X
DOI:	10.1016/j.est.2024.113619