Journal of Alternative and Renewable Energy Sources

Learning Battery Health from Degradation Indicators: A Critical Review of Data-Driven SOH Estimation

Ranvijay Parmar, Dev Karan Sakravdia — Tue, 15 Sep 2026 00:00:00 +0000

The accurate estimation of the state-of-health (SOH) of lithium-ion batteries is essential for reliable battery management, safety assessment and remaining service-life planning. Conventional SOH estimation methods often rely on electrochemical models, equivalent-circuit representations or controlled diagnostic measurements, which may require detailed battery knowledge, parameter identification or operating conditions that are difficult to maintain in practical applications. Data-driven approaches provide an alternative by learning the relationship between measurable battery signals and degradation behaviour from experimental or operational data. The reviewed approaches are analysed across conventional machine-learning methods, deep-learning architectures, hybrid and ensemble models, transfer and domain-adaptation strategies, and uncertainty-aware approaches. Particular attention is given to the role of input data, feature representation, model complexity, interpretability, computational requirements and generalisation capability. Feature-based machine-learning methods remain relevant where informative degradation indicators and limited computational resources are available, whereas deep-learning approaches provide greater capacity for automatic representation learning but require sufficiently diverse and representative datasets. Hybrid and knowledge-guided approaches offer potential advantages by combining complementary information sources, although their additional complexity must be justified through improved reliability or transferability. The review further identifies data quality, uncertain SOH labels, operating-condition variability, domain shift, feature instability, inadequate validation and deployment constraints as major barriers to practical implementation. A central finding is that reported prediction accuracy alone is insufficient for evaluating data-driven SOH estimators, as model performance is strongly influenced by dataset composition and validation design.

AI-driven Renewable Energy Generation Forecasting Using Machine Learning Techniques

Arathi H L, Sinchana C S, Partha C M, Shobha N, Shreya K S — Tue, 08 Sep 2026 00:00:00 +0000

Solar and wind power depend heavily on weather, so the amount of electricity they will supply in the next few hours is difficult to know in advance. This makes planning harder for the people who operate power systems. Machine learning is widely used to predict renewable generation, and recent studies report good prediction accuracy. However, most of these studies stop at the prediction itself. They give a single number without saying how much it may vary; they do not explain which weather conditions led to that number; they work on stored data files instead of live weather information; they are trained only once, and they end with a comparison of error values rather than with advice that an operator can act on. This paper proposes a framework for renewable energy generation forecasting that tries to solve these practical problems. The proposed system is planned as five simple layers covering data collection from stored records and a live weather service, data preparation, a set of machine learning models with a comparison step, supporting features such as a confidence range, an explanation of the influencing factors and periodic model updating, and finally a web application with a dashboard, planning suggestions and a simple assistant. The paper explains the need for the work, the gaps found in existing studies, the proposed design, how the system is expected to work, and a stage-wise plan for building and testing it. The system has not yet been developed, so no results are reported. The expected benefits are described in general terms and include more useful forecasts, better understanding of the predictions, easier use by non-technical users, and a clearer path from a research model to a working tool.

Enforcing Electrochemical Consistency: A Hard-Constrained Physics-Informed Neural Network for Battery RUL Prediction

B. Towfeeq Ahmed, A. Touseef Ahmed, P. Muzammi — Thu, 10 Sep 2026 00:00:00 +0000

Predicting the Remaining Useful Life (RUL) of lithium-ion batteries accurately is imperative in ensuring the safety and reliability of energy storage devices. Although recent physics-aware approaches based on neural networks have successfully shown the importance of extracting parameters related to battery aging from charging curves, their architecture still remains an agnostic and universal approximator without incorporating the underlying physical governing equations. In this work, they propose a Hard-Constrained Physics-Informed Neural Network (HC-PINN) where the governing differential equation of Solid Electrolyte Interphase (SEI) growth and Arrhenius temperature dependency is explicitly integrated within the loss function. Contrary to soft-constrained loss functions, which only penalize the deviation, this approach uses a hard residual loss to penalize violations of the square root time degradation law and Arrhenius law. Numerical results based on the selected battery dataset show that under optimal conditions, HC-PINN reaches an RMSE score of 9.87 cycles, significantly better than the physics-featured FCNN baseline (11.42 cycles). Importantly, when applying 15% Gaussian noise to input data (to represent sensor errors in reality), HC-PINN reaches an RMSE value of 14.23 cycles, whereas FCNN (unconstrained neural network) achieves only 38.56 cycles. Their novel approach brings a new direction in embedding the thermodynamic consistency constraint into neural network optimization.