Estimating Remaining Useful Life and State of Health of EV Lithium Ion Batteries Using Sequential CNNs
Authors:
- 8 December 2023
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-6
Abstract
Lithium-ion batteries power electric cars, portable electronics, and renewable energy systems. To ensure these batteries' optimal performance, avoid unexpected failures, and reduce maintenance costs, their usable life must be accurately predicted. This study examines data-driven, physics-based, and hybrid methods for calculating lithium-ion battery life. The proposed work also offers a unique way to estimate lithium-ion battery usable life using machine learning. This method uses current, voltage and temperature to build a prediction model that accurately predicts the battery's remaining usable life. An archive of lithium-ion battery cycles is used to evaluate the proposed strategy and compare it to other cutting-edge methods. The results support the proposed lithium-ion battery lifespan strategy. The precise estimation of battery SOH ensures uninterrupted operation and prevents unexpected battery failures, improving electric vehicle reliability and safety. It also optimizes battery use, extending battery life and lowering replacement costs. The proposed method could improve electric vehicle (EV) performance and efficacy using machine learning and electrochemical modeling, promoting sustainable transportation. Portable electric machines, electric locomotives, and renewable energy storage systems use lithium-ion batteries. For optimal performance, cost reduction, and preventing unexpected failures, accurate SOH estimation is essential. This paper examines data-driven, physics-based, and hybrid SOH estimation methods. This paper proposes a machine learning-based SOH prediction method that uses battery performance parameters like voltage, current, and temperature. The proposed work shows that this method outperforms state-of-the-art methods on a battery cycle dataset. EV reliability, safety, and battery utilization improve with accurate SOH estimation, lowering replacement costs. The proposed method can improve EV performance and promote sustainable transportation by using machine learning and electrochemical modeling.Keywords
This publication has 10 references indexed in Scilit:
- A Data-Driven-Based Framework for Battery Remaining Useful Life PredictionIEEE Access, 2023
- Stacking-based ensemble learning for remaining useful life estimationSoft Computing, 2023
- A QoE Framework for Video Services in 5G Networks with Supervised Machine Learning Approach2023
- A Novel Transfer Ensemble Learning Framework for Remaining Useful Life Prediction Under Multiple Working ConditionsIEEE Transactions on Instrumentation and Measurement, 2023
- Scalability, Explainability and Performance of Data-Driven Algorithms in Predicting the Remaining Useful Life: A Comprehensive ReviewIEEE Access, 2023
- Remaining Useful Life Prediction of Lithium-Ion Batteries Based on Data Preprocessing and Improved ELMIEEE Transactions on Instrumentation and Measurement, 2023
- Remaining Useful Life Prediction of Turbofan Engines Using CNN-LSTM-SAM ApproachIEEE Sensors Journal, 2023
- Machine Learning Based Classifiers for QoE Prediction Framework in Video Streaming over 5G Wireless NetworksComputers, Materials & Continua, 2023
- A Review of Lithium-Ion Battery State of Health Estimation and Prediction MethodsWorld Electric Vehicle Journal, 2021
- Remaining Useful Life Prediction for Complex Systems With Multiple Indicators Based on Particle Filter and Parameter CorrelationIEEE Access, 2020