Applied Sciences

Sepsis early warning is hindered by data silos, temporal leakage, and threshold choices that obscure operational performance. We present a leakage-aware federated-learning evaluation pipeline that enforces group/temporal separation and compares models at a fixed alert workload. Stage-1 benchmarks local, FedAvg, and FedProx LSTM/Transformer models on PhysioNet/CinC 2019 using the official A/B partitions in bidirectional cross-hospital evaluation (A→B/B→A) after removing ICULOS. Stage-2 constructs a Sepsis-3-aligned MIMIC-IV task using full SOFA-component features and simulated clients to emulate institutional heterogeneity. Federated training improves out-of-hospital generalization for LSTM models on PhysioNet, whereas Transformer models remain robust across 3–12 h horizons. On MIMIC-IV, fixed alert-rate evaluation (α = 5%) clarifies workload–timeliness trade-offs, and centralized XGBoost achieves the strongest stay-level detection with clinically meaningful lead times. Supplementary privacy and security stress tests further contextualize residual deployment risks. Overall, leakage control and workload-matched evaluation are essential for trustworthy, operationally actionable sepsis early warning.

Identifying real-world saturation points and grid-hosting capacity in mixed-use urban Renewable Energy Communities (RECs) requires dynamic spatial evaluation. To address this, this paper introduces a novel simulation framework that integrates GIS spatial analysis with an iterative heuristic selection algorithm. The proposed method evaluates the energetic interaction between a primary generation node and surrounding consumers, utilizing a dynamic function to calculate the collective Self-Consumption Rate (SCR). Applied to the Flaminio Stadium in Rome, the model incrementally aggregates users to determine the optimal cluster size for economic feasibility. The results demonstrate that the heuristic selection algorithm successfully refined the community from an initial pool of 854 buildings to an optimal cluster of 734. This targeted selection eliminated energy surplus and achieved a near-perfect collective SCR of 99.8%. Furthermore, by strategically reducing the required installed PV capacity by 52.6%, the initial capital investment dropped from € 89.9 million to € 42.6 million, significantly de-risking the project while maintaining a competitive payback period of approximately 13 years. Ultimately, this study presents a scalable spatial optimization tool that empowers decision makers to transform large-scale urban infrastructure into the energetic and economic engines of district wide decarbonization

Carbon fiber-reinforced polymer (CFRP) has been widely used in various fields due to its significant advantages. However, research on their pyrolysis and combustion behavior under fire conditions, which directly affects structural integrity and safety, remains insufficient. To challenge this issue, thermogravimetric analysis was employed to investigate the pyrolysis characteristics of the CFRP in both air and nitrogen atmospheres at heating rates of 20–40 °C/min with relevant pyrolysis kinetic parameters calculated using the Kissinger method. Fourier-transform infrared (FTIR) spectrometer was utilized to analyze pyrolytic gas species and concentrations at 40 °C/min in nitrogen atmosphere. Cone calorimeter tests at 50 kW/m² were conducted to obtain combustion characteristic parameters. Based on atomic conservation and oxygen-consumption principles, the equivalent molecular formula (CH_5.787O_0.541) of the epoxy resin pyrolysis gas and its combustion reaction equation were derived through reverse deduction. The heating, pyrolysis, and combustion processes of the CFRP (cone calorimetry specimen) were numerically simulated using Fire Dynamics Simulator (FDS). The predicted heat release rate, mass loss rate, and gas production rate showed good agreement with experimental results.