Sealing quality critically determines the corrosion resistance of anodic oxide films on aluminum alloys. This study systematically investigated lithium nitrate-based sealing strategies, including single-step treatments at 50, 75, and 90 °C, and stepwise approaches with varying temperature gradients (50–75 °C, 50–90 °C, 50–75–90 °C). The results reveal that sealing strategies significantly impact sealing product composition and spatial distribution, thereby influencing film corrosion resistance. Low-temperature (50 °C) single-step sealing yielded boehmite and amorphous hydroxides in inner film regions, while higher temperatures (75/90 °C) promoted rapid formation of boehmite and Li–Al-LDH (lithium–aluminum layered double hydroxide) in outer regions. Stepwise sealing combines deep low-temperature penetration with high-temperature Li–Al-LDH formation kinetics, creating gradient composite structures that enhance film compactness and stress relief. Electrochemical and salt spray tests demonstrated that 50–75 °C stepwise-sealed samples exhibited the lowest corrosion current density and highest impedance, with superior performance in both neutral (pH 7.0) and acidic (pH 3.5) environments. This work elucidates lithium salt sealing mechanisms through temperature control and provides guidelines for eco-friendly, high-performance aluminum surface treatments.
The development of science and technology has promoted the unmanned aerial vehicle (UAV) industry. Due to its small size, lightweight, low cost, and other characteristics, UAV can integrate with multiple industries, and promote social development, which broadens the use of UAV itself. UAVs have been widely used in aerial photography, agriculture, and disaster rescue. This paper analyzed the application of UAV in geological disaster rescue. Using UAV remote sensing to photograph the roads to the geological disaster area, the road conditions of different roads could be analyzed, providing the best rescue route in the disaster area. The current point-feature-based methods fail to accurately identify and analyze the target in the UAV image. This paper proposed a convolution neural network (CNN) based model to analyze the UAV image by automatically identifying the image targets. We investigated the accuracy of vehicle recognition using traditional UAV image recognition and our CNN-based model. The experimental results showed that the proposed method improved the average recognition accuracy by 9.35 and 9.08
Different from previous studies, this paper is motivated to examine supply chain risk from a complex network perspective rather than a simple supply chain structure, and considers risk propagation and intervention in a complex supply chain network (CSCN) under public emergency. Facing the crisis triggered by public emergency, the government decides to aid partial enterprises in order to improve the stability of the CSCN. Inspired by the similarity between the risk propagation in the CSCN and the diffusion of infectious disease in the social network, this study tries to develop a new risk propagation model of the CSCN based on the SEIR system dynamics theory in infectious disease field. Using the classic basic reproduction number (BRN) concept, which means the average number of enterprises that one infected enterprise will infect, this study examines the risk propagation behavior and its threshold in the CSCN. This study finds out that the government can effectively control the risk propagation in the CSCN by way of controlling the BRN. Under the control of government, if the BRN is smaller than one, then the risk will withdraw. On the contrary, if the BRN is bigger than one, then the risk will diffuse to the entire supply chain network at all. This study also finds out that the risk propagation in the CSCN is relatively sensitive to the recovery probability. This study further conducts simulation analysis of risk propagation, the results demonstrate that the risk will be controlled or abated by controlling the BRN.
Flavonoids with photoluminescence were widely recognized as naturally occurring polyphenolic compounds, celebrated for their potent antioxidant properties. The antioxidant efficacy of these flavonoids can be precisely tuned by structural modification, a finding with profound implications for biomedicine, particularly in cancer diagnosis and anticancer drug development. However, despite their promising prospects, the clinical translation of flavonoids is hampered by their suboptimal aqueous solubility, low gastrointestinal bioavailability, rapid systemic clearance, and limited targeted delivery capabilities. Fortunately, glycosylation offers a promising strategy to ameliorate these constraints, thereby potentiating their pharmacodynamic activity. In the present investigation, density functional theory (DFT) and time-dependent DFT (TD-DFT) were utilized. A comprehensive evaluation of the antioxidant properties was conducted for baicalein exhibiting excited-state intramolecular proton transfer (ESIPT) with a non-existent enol⁎ state fluorescence and 6,7,4'-trihydroxyisoflavone without ESIPT characteristics, both in aqueous solution. Research indicates that the antioxidant activity of flavonoids is significantly enhanced in their excited states. Attributed to their ESIPT properties, baicalein and its derivatives exhibit superior antioxidant performance in their keto* state. Additionally, glycosylation was found to contribute to the further enhancement of the flavonoid's antioxidant activity. These findings underscore the pivotal role of ESIPT and glycosylation in the design of high-efficacy flavonoid-based antioxidants.
As artificial intelligence (AI) advances, product design urgently requires frameworks to accurately interpret user needs and generate data-driven design schemes. This study proposes a generative artificial intelligence (GenAI) methodology for furniture design that integrates semantic mining, prompt engineering, and multi-criteria evaluation. A hybrid strategy combining Latent Dirichlet Allocation (LDA) and grounded theory is employed to extract user needs from 75,895 online reviews, thereby establishing a hierarchical semantic framework comprising five core dimensions and nineteen sub-criteria. The Fuzzy Analytic Hierarchy Process (FAHP) is applied to quantify the relative importance of user needs. Weighted text prompts are then generated using large language models (LLMs). These optimized prompts are fed into a Stable Diffusion model (SD) enhanced with Low-Rank Adaptation (LoRA) and guided by ControlNet constraints. This generates visual prototypes aligned with user semantics. For final evaluation, the framework uses the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) to rank the generated schemes across the nineteen identified criteria. A case study involving ergonomic office chairs quantitatively demonstrates this method’s effectiveness in improving user preference interpretation, design quality, and decision-making transparency. By bringing together user need identification, generative concept generation, and objective evaluation, this end-to-end AI framework provides a theoretically grounded approach to advancing the intelligent transformation of industrial design.
