In order to reveal the macroscopic thermodynamic characteristics and microscopic reaction mechanism of low-temperature oxidation of lignite in eastern Inner Mongolia, this study systematically studied it by combining thermogravimetric analysis (TG), programmed temperature experiments, and ReaxFF MD molecular dynamics simulations. The experimental results indicate that the apparent activation energy (Ea) exhibits a dynamic pattern of decrease increase decrease. During the dehydration stage, Ea decreases with an average value of 46.29 kJ/mol; During the oxygen adsorption stage, the apparent Ea increased to 373 kJ/mol, reflecting the enhanced effective kinetic resistance caused by desorption adsorption competition and intermediate accumulation; Then during the combustion phase, with the intensification of structural activation and heat accumulation, Ea decreases to 60.25 kJ/mol. The ReaxFF MD simulation captured a phased mass evolution consistent with the TG results and further revealed the atomic level formation pathways of key gaseous products. CO is mainly related to the cleavage pathway of ketones/aldehydes, which is consistent with the energy barriers calculated by density functional theory (DFT) previously reported. CO2 mainly comes from carboxyl decomposition and secondary CO oxidation, while the formation of H2O is driven by hydrogen extraction involving hydroxyl radicals. C2H4 is mainly generated through the cleavage and recombination of long-chain aliphatic side chains. By combining the kinetic parameters obtained from TG, programmed temperature gas release, and ReaxFF MD reaction trajectory, this study established a connection between macroscopic thermal behavior and microscopic bond breaking mechanisms, providing a theoretical basis for early warning and prevention of low rank coal spontaneous combustion.
The tunnel convergence pattern, a key boundary condition in analyzing strata deformation induced by tunnel excavation, is influenced by complex geological conditions and construction activities. However, existing analytical approaches rely on idealized convergence assumptions, limiting their ability to accurately reflect strata deformation under realistic engineering conditions. Building upon a previously developed dual-engine predictive framework, this study refines the analytical formulation and inversion algorithm while incorporating the complexity of convergence patterns, thereby enabling more efficient and refined prediction of strata deformation. A unified equation for tunnel convergence patterns is formulated based on Fourier series and coordinate transformations, enabling an accurate description of the tunnel contour trajectory. Using the transformation relationships among three independent coordinate systems, an analytical relationship between the unified equation parameters and tunnel contour deformations are established, and a complex variable solution for strata deformation under arbitrary convergence patterns is derived. Model degeneration and parameter sensitivity analyses demonstrate that the proposed solution is a generalized form of traditional complex variable solutions, providing a more refined characterization of strata deformation. By further integrating deep learning models with strong nonlinear fitting capabilities, the inversion of convergence patterns for sample sections is achieved, and an improved dual-engine predictive approach that accounts for the complexity of convergence patterns is developed. Engineering applications demonstrate that the improved method can achieve satisfactory predictive performance, with predicted ground settlements exhibiting closer agreement with field measurements than predictions from conventional approaches. The proposed method enables efficient, real-time, and refined prediction of strata deformation, demonstrating considerable potential for engineering applications.
Fairness in same-day delivery (SDD) becomes increasingly important as customers demand equitable service pledge, yet cost-driven merchants’ preferences for nearby customers create service disparities across regions. Drone fleet resupply presents a fast and traffic-free solution capable of navigating complex terrains, while truck-based delivery complements it by overcoming the drone’s limitations in endurance and operational range. Previous studies mainly address single-drone, single-truck problems, with limited attention to the potential of drone fleet resupply with multi-truck systems to solve fairness issues. This research fills this gap by originally proposing a drone fleet resupply system for stochastic requests, where drones perform multiple trips to replenish any truck as needed, with consideration of drone endurance limitations. This problem is formulated as a sequential decision process (SDP) and solved through a deep Q-learning (DQL) approach to maximize service rates while ensuring fairness. A piecewise linear reward function is designed to improve convergence. Comparative results against benchmark policies demonstrate DQL’s potential to manage highly stochastic logistics operations. Key findings include: (1) Drone fleet resupply enhances regional fairness and overall service rates, demonstrating strong robustness in serving long-distance and time-sensitive deliveries; (2) An appropriate reward function significantly facilitates fairness, while penalties further improve it but are ineffective alone; (3) An optimal drone-truck fleet mix exists to balance service rates and associated costs, while adding more trucks alone substantially improves service rates but incurs larger costs; (4) The proposed approach remains effective across drone capacities, while extended drone endurance yields limited gains in service rate but significantly lowers delivery costs.
Retrofits for ultra-low emissions couple residual energy recovery with pollution control because treatment units alter flue gas states and downstream recovery pathways. This study develops an exergy–cost optimization framework that jointly configures flue gas treatment, by-product gas conversion, multi-grade waste heat recovery, terminal energy services, and in-plant transport. Mechanism-constrained grey-box modules propagate stream states through reconfigurable treatment chains, and NSGA-II searches Pareto solutions under emission, operating-window, acid-dew-point safeguard, sulfur-tolerance, capacity, and topology constraints. For the synthetic reference steelworks, recoverable source exergy was 341.13 MW, 72.2% of which originated from by-product gases. The balanced solution selected an ESP–WFGD–reheater–WHPG–SCR route, with 238.06 MW augmented system exergy loss, a 30.21% recoverable source exergy utilization index, a 6.29% unused-source ratio, and a net system cost of − 57.63 × 103 CNY/h. Across twelve scenarios, pollutant loading altered treatment-chain selection, gas availability strongly affected power generation and economic response, and seasonal demand redistributed low-grade heat. Removing the unused source opportunity term increased unused recoverable exergy by 67.97 MW. Under high NOx loading, 10% tighter pollutant set points removed the WHPG branch with only a 0.038% increase in augmented loss. Plant application requires site-specific recalibration and independent validation.
Direct regeneration of spent LiFePO4 (LFP) cathodes faces two critical challenges: inhomogeneous contact between repair agents and degraded particles, and insufficient cycling stability of regenerated materials. Here, we address both issues through a two-step powder-level regeneration strategy combining high-speed slurry homogenization with fluorine-source treatment. The spent LFP powder, comprising mixed LiFePO4 and FePO4 phases, is fully restored to phase-pure olivine LiFePO4 after thermal annealing. Replacing conventional stirring with high-speed homogenization significantly improves the distribution of LiOH·H2O and melamine precursors around LFP particles, boosting the initial discharge capacity from 135.88 to 152.38 mAh·g−1 at 0.1C. Subsequent fluorine treatment using LiF or NH4F introduces uniformly distributed fluorine species on particle surfaces, as confirmed by SEM-EDS mapping. Although fluorination slightly reduces the initial capacity (to ∼142 mAh·g−1), it substantially enhances long-term cycling stability, with capacity retention at 1C after 400 cycles increasing from 87.74% for the non-fluorinated sample to 93.85% for LiF-RLFP and 92.63% for NH4F-RLFP. The fluorine-treated materials retain ∼90% capacity after 100 cycles in LFP||graphite full cells, while further analyses indicate improved kinetics and reduced accumulation of electrolyte-derived interfacial species. This work offered a simple and scalable route for spent LFP direct regeneration.
