The rapid development of the digital economy has led to a significant increase in the scale and electricity load of 5G base stations in urban areas. It is therefore essential to study the load characteristics of 5G base stations to enhance urban grid stability and improve the energy efficiency of telecom operators. This paper investigates the load characteristics of 5G base stations using energy data from 37,525 5G base stations. The comprehensive analysis examines spatial distribution patterns, temporal load characteristics, and grid integration impacts using 15-minute resolution data. Results show that base station deployment is strongly correlated with population density (r = 0.92). Temporal load patterns show strong seasonal characteristics, with summer consumption about 1 kW higher than winter consumption, driven primarily by temperature-dependent cooling demand. The base stations maintain consistently low capacity utilization rates (26-41%), indicating substantial infrastructure redundancy. Strong synchronization exists between base station and grid loads during peak consumption periods, increasing peak stress but also revealing significant demand response potential. This study can help urban energy facilities identify new flexible sources and improve 5G station planning and operations.

Ultra-high performance concrete (UHPC), an advanced cementitious composite characterized by superior mechanical properties and durability, requires multi-objective collaborative optimization in mix proportion design to facilitate its large-scale engineering applications. To address this challenge, this study proposes an intelligent design framework for UHPC driven by the synergistic integration of Stacking ensemble learning and the non-dominated sorting genetic algorithm III (NSGA-III). In the model construction stage, a Stacking ensemble model was developed by integrating eight heterogeneous algorithms as base learners and employing linear regression as the meta-learner. Subsequently, six hyperparameter optimization strategies were employed to fine-tune the model. At the optimization decision stage, the optimal Stacking ensemble model was embedded into the NSGA-III algorithm, and the technique for order preference by similarity to ideal solution (TOPSIS) was employed to select the optimal mix design from the Pareto front. The results demonstrate that the predictive accuracy of the Stacking ensemble model significantly outperforms that of the individual single-algorithm models. Furthermore, multi-objective optimization conducted for three target strength grades, UC100, UC120, and UC140, revealed that all optimized schemes successfully meet the specified strength requirements while effectively balancing environmental impact and production cost. This research establishes a data-driven, multi-criteria intelligent decision-making pathway for the multi-objective design of UHPC, with substantial implications for advancing the interdisciplinary integration of materials science and artificial intelligence.

Ultraviolet germicidal irradiation (UVGI) is widely applied in heating, ventilation, and air-conditioning systems to reduce airborne pathogen transmission, yet its effectiveness depends on airflow-driven changes in lamp thermal condition and particle residence time. This study experimentally characterises the coupled thermal and optical performance of 95 W and 60 W low-pressure mercury ultraviolet-C (UV-C) lamps and uses the measured irradiance fields for comparative germicidal assessment. Air velocities of 1.1-2.5 m/s and ambient temperatures of 14-22 °C were varied to quantify their effects on lamp-surface temperature and 254 nm output. The 95 W lamp maintained near-maximum output over a broader operating range, whereas the 60 W lamp was more sensitive to convective cooling. Reynolds and Rayleigh numbers ranged from approximately 1.24 × 10³ to 2.88 × 10³ and 7.2 × 10³ to 1.3 × 10⁴, respectively, while the convection-to-radiation heat-loss ratio increased with airflow velocity. Mean modelled ultraviolet (UV) dose increased with lamp count and decreased with airflow velocity. At 1.1 m/s, six-lamp arrays delivered mean modelled doses of approximately 191 J/m² for the 95 W lamps and 146 J/m² for the 60 W lamps. Published UV susceptibility constants produced calculated reductions exceeding 6 log for susceptible viral and bacterial species, whereas the calculated reductions for Aspergillus spores ranged from below 0.05 to approximately 0.4 log. These microbial reductions were calculated from the measured irradiance fields and published susceptibility constants and were not experimentally validated using bioaerosols. The results provide a measurement-informed basis for comparing lamp power, airflow condition, and lamp arrangement in in-duct UVGI systems.

The rapid development of the low-altitude economy has positioned vertiports as critical heavy-asset hubs in urban logistics networks, where siting decisions affect both delivery access and facility utilization. Existing maximum-coverage models do not explicitly represent station throughput or allocated order volumes. This study therefore retains the original three-stage framework-demand assessment, hierarchical facility polygon intersection point set (HFPIPS) candidate generation, and capacitated location-allocation, while correcting the demand surface to use area-weighted land-use shares and clarifying that the utilization floor is an application-specific minimum-throughput rule rather than a new algebraic form. In Nanjing, 8,811 demand grids and 10,283 planning candidates produce 2,678,898 radius-feasible pairs, reduced to 374,873 pairs by demand-adaptive pruning. Under a 300 s single-thread budget, the high-coverage capacitated facility location problem (CFLP) returns a verified feasible incumbent of 195 stations delivering 156,300 orders/day (78.13%), with 66.79% average utilization, 22.50% minimum utilization, and 0.953 km average delivery distance. At the same 195-site budget, a deterministic maximal covering location problem (MCLP)-greedy baseline achieves 97.05% nominal geometric coverage but only 63.87% capacity-feasible delivery. Removing the utilization floor delivers only 29 additional orders while opening four extra sites and permitting utilization as low as 1.58%. Land-use coefficient contrast tests change delivery by less than 0.27 percentage points and preserve the selected station set. The results support capacity-aware planning, while building-level engineering data, heterogeneous fleets, temporal demand, and unresolved solver gaps remain limitations.

This study develops a life-cycle techno-economic analysis and optimization framework for hybrid deep ground source heat pump (GSHP) systems, addressing the limitations of existing evaluations that often neglect subsurface thermal attenuation and rely on simplified heat transfer models. A coupled heat transfer model is established by combining the finite difference method for the borehole interior with the segmented finite line-source model for the surrounding ground, enabling long-term simulation of fluid and ground temperature evolution. The energy model is integrated with a life-cycle economic evaluation to calculate net present value, levelized cost of heating/cooling, payback period (PP), and internal rate of return. A Bayesian optimization algorithm is further employed to identify optimal system configurations under multiple objectives. The framework is applied to a hybrid geothermal system in Xi’an, China, incorporating mid-deep boreholes with auxiliary air-source heat pumps or gas boilers. Results show that long-term thermal attenuation significantly affects economic performance, with a 2 °C increase in fluid temperature extending the PP by approximately 0.3 years. User-side load uncertainty also has a substantial impact, as a 10% deviation from the design load can shift the optimal PP by about 2 years. Sensitivity analysis identifies the GSHP installed capacity ratio and borehole number as the most influential parameters. Compared with pure geothermal systems, the hybrid geothermal–gas boiler configuration achieves the shortest PP, demonstrating the framework’s effectiveness for robust low-carbon system design.

Alkali-activated lunar regolith is considered one of the most promising lunar regolith-based construction materials for large-scale lunar construction. This study presents a comprehensive framework integrating machine learning (ML) algorithms to investigate the influence of various features on the mechanical strength of alkali-activated lunar regolith simulant (AALRS), aiming to achieve the strength prediction and optimization design of AALRS. The properties of lunar regolith simulant, mixture proportions, preparation parameters, environmental conditions, and enhancement methods were employed as input features for ML modeling. The compressive and flexural strength predictive models were constructed using eight ML algorithms and evaluated through statistical indicators. Among the models, Extreme Gradient Boosting demonstrated the best performance, yielding an R² of 0.8684, a root mean square error of 6.2007 MPa, and a mean absolute error of 4.0874 MPa on the testing dataset. Using the best prediction models, four design strategies with three objectives, namely, compressive strength, flexural strength, and transport payload (TP), were optimized and evaluated using the non-dominated sorting genetic algorithm II and the technique for order preference by similarity to ideal solution methods. The proposed prediction and optimization framework for the mechanical performance of AALRS, which integrates extreme environmental effects and TP-driven design, provides a robust data-driven approach for the design, prediction, and optimization of AALRS, advancing the development of extraterrestrial construction materials and technologies.