Toward accurate remaining useful life and state of health estimation using reinforced graph-based physics-informed neural networks enhanced with dynamic weights

Accurate estimation of Remaining Useful Life (RUL) and State of Health (SoH) is essential for reliable Prognostics and Health Management (PHM), supporting timely maintenance and dependable industrial operation. However, hybrid models that combine data-driven learning with physics-based regularization often rely on fixed loss weights and can be difficult to calibrate when separately applied to assets with different degradation behaviors. This study introduces Reinforced Graph based Physics-informed Networks with Dynamic Weighting (RGPD), a unified framework for spatio-temporal degradation modeling and adaptive physics guided regularization. Graph-based representation learning captures attention-weighted temporal relations among multi-sensor system states, a Soft Actor-Critic (SAC) module refines latent features under noisy conditions, and a lightweight Q-learning policy adaptively balances monotonicity, smoothness, and latent-dynamics residual losses during training. The framework is evaluated through separate instantiations on the C-MAPSS, PHM2012, and XJTU benchmarks, representing engine, bearing, and battery degradation processes. Relative to the strongest compared baselines reported in the corresponding benchmark tables, RGPD improves average RMSE by up to 12% on PHM2012 and C-MAPSS, and reduces average MAPE by 20% on XJTU compared with the second-best reported model. Performance across these separately configured heterogeneous benchmarks demonstrates methodological generality of the framework design. The physics-informed component combines degradation-consistent priors with a Deep Hidden Physics Model-style residual to promote physically plausible degradation trajectories across asset types.