Open Access

ARTICLE

Lifetime Prediction of LiFePO4 Batteries Using Multilayer Classical-Quantum Hybrid Classifier

Muhammad Haris^1,*, Muhammad Noman Hasan¹ , Abdul Basit², Shiyin Qin¹

1 School of Automation Science and Electrical Engineering, Beihang University, Beijing, 100083, China
2 China Academy of Space Technology, Shenzhou Institute, Beijing, 100010, China

* Corresponding Author: Muhammad Haris. Email:

Journal of Quantum Computing 2021, 3(3), 89-95. https://doi.org/10.32604/jqc.2021.016390

Received 02 May 2021; Accepted 10 August 2021; Issue published 21 December 2021

Abstract

This article presents a multilayer hybrid classical-quantum classifier for predicting the lifetime of LiFePO4 batteries using early degradation data. The multilayer approach uses multiple variational quantum circuits in cascade, which allows more parameters to be used as weights in a single run hence increasing accuracy and provides faster cost function convergence for the optimizer. The proposed classifier predicts with an accuracy of 92.8% using data of the first four cycles. The effectiveness of the hybrid classifier is also presented by validating the performance using untrained data with an accuracy of 84%. We also demonstrate that the proposed classifier outperforms traditional machine learning algorithms in classification accuracy. In this paper, we show the application of quantum machine learning in solving a practical problem. This study will help researchers to apply quantum machine learning algorithms to more complex real-world applications, and reducing the gap between quantum and classical computing.

---

Keywords

---