天津城建大学

District heating system (DHS) operation requires allocating heating loads among buildings with heterogeneous thermal responses while maintaining indoor comfort and avoiding excessive heat supply, particularly under volatile weather and capacity constraints. We propose a spatiotemporal forecasting-based multi-objective reinforcement learning (STF-MORL) framework that leverages building-level differences in thermal response to guide adaptive heating load allocation. The framework first uses building-specific autoregressive with exogenous inputs (ARX) models to derive coefficient-based empirical control features and generate look-ahead indoor-temperature trajectories. These features are then embedded in the agent state and reward formulation, while Pareto solution sets generated using the non-dominated sorting genetic algorithm II (NSGA-II), together with adaptive objective weighting, are used to balance fairness, comfort, and efficiency under extreme operating conditions. Using measured data from five residential buildings in Tianjin, China, STF-MORL reduces inter-building temperature dispersion by 11.7–13.8 % relative to proportional allocation across four extreme operating conditions while maintaining 100 % compliance with the 18-22℃ comfort range. It also increases the heating-load saving index by 2.7–6.2 percentage points relative to Proportional Allocation. These results provide a case-study demonstration of how building thermal heterogeneity can be translated into actionable heating-load allocation policies.

The effects of reaction temperature and calcium oxide addition on the content and forms of nitrogen and phosphorus in the thermal solvent extraction products of municipal sludge were systematically investigated. In addition, the LCA method was introduced for the first time into this process. Three quantitative indicators including nitrogen extraction efficiency (NEE), phosphorus enrichment factor (PEF), and phosphorus bioavailability conversion ratio (AP/TP) were constructed. By combining experimental data with the Ecoinvent database, the carbon emissions, net energy consumption, and substitution emission reduction benefits of nitrogen and phosphorus recovery under different operating conditions were quantified. The optimal conditions were identified as 300 °C with a CaO dosage of 4 g/100 g sludge, under which both NEE and PEF reached their maximum values. Below this temperature, the thermal solvent extraction reaction was dominated by pyrolysis, whereas at higher temperatures, decomposition and condensation were prone to occur, leading to the formation of polycyclic aromatic hydrocarbons or graphitized structures. With increasing temperature and CaO dosage, the conversion of protein nitrogen to pyridinic nitrogen, pyrrolic nitrogen, and quaternary ammonium compounds was promoted, and these nitrogen species possess enhanced thermal stability. The proportion of apatite phosphorus in the residue increased from 44.6% to 72.6%. A net carbon emission of 389.5 kg CO2-eq/t dry sludge was obtained, which is considerably lower than that of direct sludge incineration.

Abstract To investigate the evolution of mechanical behavior in concrete box girder bridges during cantilever construction under the coupled influence of stochastic environmental temperature and humidity, a novel simulation method is proposed. This method integrates a temperature–humidity coupling analysis method for the internal fields of concrete and a constitutive model that accounts for multifactor interactions, including temperature-induced stress, drying shrinkage, creep, and the sensitivity of elastic modulus. The proposed method was applied to simulate and analyze the construction process of the Yuanhe Bridge in Jiangxi Province, China. Based on the statistical analysis of nearly two decades of daily historical meteorological data from the region, it was verified through fitting tests and statistical assessments that the probability distributions of temperature, humidity, and wind speed conform to Gaussian (normal) distributions. The results systematically reveal the statistical influences of stochastic environmental factors on the bridge’s camber, shear-lag effect, and prestress loss. The results indicate that the maximum camber and the shear-lag coefficient at the bottom slab–web junction exhibit slightly asymmetric distributions and can be well described by Bigaussian fitting curves, whereas the shear-lag coefficient at the top slab–web junction follows a Gaussian distribution. Increases in the daily temperature range and decreases in relative humidity are found to amplify the camber and the prestress loss. No discernible influence of wind speed on the camber, shear-lag effect, or prestress loss is observed. This study provides a crucial theoretical foundation and technical support for the refined control and optimization of cantilever construction of bridges under stochastic environmental conditions.

Abstract Considering that current research on biocorrosive building materials is predominantly conducted through field exposure in marine environments or static laboratory cultivation, field tests struggle to achieve precise quantification of individual biological species’ effects on concrete. Meanwhile, experimental studies on the impact of cultured marine fouling organisms on concrete remain in the nascent phase of static cultivation. To address these challenges and better simulate the natural marine environment, this study incorporates a wave pump within an incubator to create dynamic flow conditions. We investigate the evolution of biofilm adhesion on mortar surfaces under this dynamic environment and analyze its effects on the structural integrity and durability of the mortar. Experimental results demonstrated that flow conditions enhance the photosynthetic and metabolic activities of the U l v a - diatom biofilm organisms, significantly influencing the pH and dissolved oxygen levels of the water. Furthermore, the adhesion of the U l v a - diatom biofilm in the dynamic environment detrimentally affects the mortar’s mechanical properties: compressive strength decreased by 45.0% and flexural strength decreased by 46.6% over the 360-day period; in contrast, the control group without algae showed a decrease of 34.1% in compressive strength and 39.0% in flexural strength over the same period. Additionally, the calcium carbonate ( CaCO 3 ) content within the mortar exposed to dynamic seawater initially increased but subsequently decreased due to dissolution. These microstructural changes also impacted the mortar’s electrochemical impedance, resulting in a 60.60% decrease in electrical resistance over 360 days; in contrast, the control group exhibited only a 36.49% decrease in electrical resistance over the same 360-day period.

Abstract Due to its high brittleness and low tensile strength, concrete is vulnerable to cracks, which will compromise the safety and service life of a structure. Although microbial repair materials have gained attention for their compatibility with concretes and their eco-friendly nature, they have limitations, such as a restricted repair depth and width, and the inability to repair complex or small cracks. In this work, we prepare a magnetic microbial repair material consisting of two components. Component A comprises a magnetic microorganism–sodium alginate (SA) composite, which can flow under a magnetic field, and Component B is composed of calcium acetate and urea. By optimizing the concentrations of the comprising materials, an effective and efficient healing of penetrating cracks is achieved. The results show that the magnetic microbial repair material exhibited the best healing effect for the smallest cracks studied in this work, which were 20 mm long, 5 mm wide, and 100 mm deep. After healing for 7 days, the water absorption rate decreased by 20.8%, and the compressive strength recovery rate was as high as 99.7%, confirming the applicability of this magnetic microbial repair material in healing small, deep cracks.