29 May 2024 in Expert Systems with Applications
12 March 2025 in IEEE Transactions on Industrial Electronics
Accurately estimating the state of health (SOH) of satellite batteries is crucial for reliable satellite operation and safe decision-making. However, the complex degradation processes of batteries in space, coupled with limited labeled data, pose challenges for existing SOH estimation methods. We propose a physics-informed neural network (PINN) for satellite battery SOH estimation using only an extreme segment data within 2 min. To address battery degradation complexity, we model degradation attributes from an empirical and state-space perspective. We combine neural networks with physical principles to capture the degradation dynamics of batteries, enabling accurate SOH estimation. To tackle limited and inconsistent data acquisition per cycle, health features are constructed from relaxation voltage data within 2 min, ensuring independence from charging/discharging strategies. Unique factors specific to satellite operations are also considered. To validate the proposed method, the battery degradation experiments simulating a geostationary earth orbit (GEO) satellite are conducted, and an aging dataset is generated. The proposed method achieves a mean absolute percentage error (MAPE) of 1.33% for SOH estimation using a 2-min extreme data segment. Our studies provide valuable insights into battery prognosis and health management for space missions, aiding in optimizing battery performance and longevity in satellite operations.
19 June 2025 in IEEE Internet of Things Journal
Accurately predicting the state of health (SOH) of lithium-ion batteries is crucial for optimizing usage and extending battery lifespan. In the era of the Internet of Things (IoT), collaborative modeling represents a pivotal trend in future SOH estimation research. However, privacy constraints induce data isolation among battery organizations, limiting collaborative progress. Federated learning (FL) enables multiparty collaboration while preserving data privacy, but it faces the inherent challenge of data heterogeneity. To address this, we propose a novel personalized federated transfer learning (PFTL) framework to enable secure collaboration and build a reliable SOH estimation model in data heterogeneity scenarios. We develop the Residual Adaptive Kolmogorov-Arnold Network (RAKAN) as the predictive model within the framework, improving the base performance of local client models. Our framework constructs global synthetic data as a core element and achieves effective aggregation among local models through the global synthetic data-guided cross-client alignment mechanism. Additionally, we design a dynamic weighted aggregation strategy based on feature-focus similarity, allowing the global model to learn more generalized feature-focus patterns. This framework can further refine the global model to obtain personalized SOH estimation models tailored to the characteristics of each client. Experimental results confirm that our method achieves robust and accurate SOH estimation results while ensuring data privacy.
8 March 2026 in Pattern Recognition