Title: Predicting battery remaining useful life based on opposition-based flower pollination algorithm and XGBoost with optimised boosting hyperparameters

Authors: Bighnaraj Naik; Rajat Kumar Sahu; Geetanjali Bhoi; Manohar Mishra; Jai Govind Singh

Addresses: Department of Computer Science and Engineering, Veer Surendra Sai University of Technology (VSSUT), Burla, Sambalpur, Odisha, 768018, India ' Department of Computer Science and Engineering, Veer Surendra Sai University of Technology (VSSUT), Burla, Sambalpur, Odisha, 768018, India ' Department of Computer Science and Engineering, Veer Surendra Sai University of Technology (VSSUT), Burla, Sambalpur, Odisha, 768018, India ' Department of Electrical and Electronics Engineering, Institute of Technical Education and Research, Siksha O Anusandhan University, Bhubaneswar, 751030, India ' Faculty of Climate Change and Sustainability, AIT Thailand

Abstract: In order to assure safety, economical, and reliability in energy storage mechanism, the accurate prediction of battery remaining useful life (RUL) is essential, particularly in battery operated portable devices and electric vehicles. This paper presents a framework that employs the extreme gradient boosting (XGBoost) model for accurate RUL prediction. XGBoost has the ability to address nonlinear relationships, missing values, and complex interactions of features render it a suitable option for predicting battery degradation behaviour modelling. However, predictive performance of XGBoost is extremely sensitive to its hyperparameter selection. To address this issue, the paper proposes the opposition-based flower pollination algorithm (OFPA) with opposition-based learning mechanism for optimising hyperparameter. OFPA is used to optimise XGBoost boosting hyperparameters such as learning rate, maximum depth, regularisation coefficients, and child weight. XGBoost-OFPA approach shows enhanced performance in predictive accuracy and computational efficiency compared to other metaheuristic methods. This study may provide a framework for smart battery health prognosis.

Keywords: battery remaining useful life; RUL; XGBoost; hyperparameter optimisation; flower pollination algorithm; FPA; opposition-based learning; OBL; predictive maintenance; battery health prognostics.

DOI: 10.1504/IJAMECHS.2026.155367

International Journal of Advanced Mechatronic Systems, 2026 Vol.13 No.3, pp.202 - 217

Received: 30 Jul 2025
Accepted: 10 Mar 2026
Published online: 30 Jul 2026 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article