Anhui Sanlian University

Major depressive disorder (MDD) pathogenesis is associated with immune-inflammatory dysregulation. This study investigated the mechanistic role of caspase-1-mediated inflammatory pathways in MDD, exploring their therapeutic targeting potential and clinical predictive utility. Peripheral blood levels of caspase-1, IL-1β, and IL-10 were assessed in MDD patients, with longitudinal analysis of dynamic changes pre-/ post-antidepressant treatment. The predictive performance of the biomarker was validated using leave-one-out cross-validation (LOOCV) and receiver operating characteristic (ROC) analysis. A chronic unpredictable stress (CUS) rat model was established to evaluate the combined therapeutic effects of caspase-1 inhibitors with fluoxetine. Clinical data revealed that expression levels of caspase-1 and IL-1β exhibited an negative correlation with clinical treatment response in patients with MDD, with varying levels observed among patients with differing depression severity. In contrast, IL-10 expression demonstrated a positive correlation with therapeutic efficacy. Following treatment, significant reductions in caspase-1 and IL-1β levels were observed, concurrent with an elevation in IL-10. Caspase-1 demonstrated superior predictive accuracy for treatment response (0.8529 ± 0.3542), yielding an area under the ROC curve (AUC) of 0.72. In animal models, co-administration of a caspase-1 inhibitor with fluoxetine accelerated the onset of antidepressant effects, while simultaneously reducing peripheral blood caspase-1 and IL-1β levels. The caspase-1/IL-1β axis contributes to MDD pathogenesis by disrupting the pro-inflammatory/anti-inflammatory balance. Its mediated immune response pattern represents a potential therapeutic target for novel antidepressant development. Combined caspase-1 modeling provides a reliable tool for predicting clinical efficacy in MDD, while caspase-1-targeted intervention strategies may overcome the limitation of slow onset of action associated with conventional therapies.

UAV technology has shown significant value in the field of highway safety monitoring. With the surge of traffic flow, traditional monitoring means are difficult to cope with the complex security management challenges, and UAVs provide innovative solutions to improve monitoring efficiency and security by virtue of high mobility, long endurance, highdefinition image acquisition and intelligent analysis capabilities. By analyzing the domestic and international application examples (such as Erguang Expressway “drone station” anti-icing to protect smoothness, Renxin Expressway Beidou drone inspection, Shaoxing Expressway “Iron Wing Warrior Team” police fast), the drone in the rapid response to accidents, abnormal events active discovery, secondary accident risk reduction and optimization of road traffic efficiency effectiveness. Reducing and optimizing the efficiency of road traffic is very effective, and it can efficiently carry out daily inspection, traffic monitoring and emergency response tasks. However, the technology still faces challenges such as insufficient endurance, weak adaptability to bad weather, data processing bottlenecks and lagging regulations. Targeted countermeasures such as optimizing battery technology, enhancing environment-adaptive design, building an intelligent data processing platform and improving regulations are proposed to promote the standardization and scale development of the technology. The study shows that the in-depth integration of UAVs and intelligent transportation system can not only significantly improve the intelligent level of highway safety management, but also provide key technical support for the construction of an efficient, safe and sustainable modern transportation system.

To mitigate the "black-and-white tunnel" phenomenon and associated traffic safety hazards at highway tunnel entrances, this study proposes an intelligent control system integrating PLC and anti-halo technology. The research begins with illumination data collection and analysis at tunnel entrances, confirming the existence of the "black-and-white tunnel" effect. Specifically, daytime tunnel entrances exhibit a sharp increase in illumination (black hole effect), while nighttime entrances show a dramatic decrease (white hole effect). Both conditions significantly impair driver visibility and pose substantial safety risks. The implementation of the PLC and anti-halo intelligent control system effectively mitigates these illumination variations during both day and night, thereby reducing visual strain on drivers. This validation demonstrates the effectiveness of the proposed solution, which holds significant practical implications for road safety.

In this study, the JPM-1 contact fatigue testing machine is used to carry out experiments on the mechanism of surface damage induced by abrasive under heavy-load line contact rolling-sliding conditions. In order to be close to the real working conditions, the lubricating oil circuit circulation system is specially designed to ensure that abrasive particles with specific concentration and particle size can continuously circulate in the experiment, so as to simulate the influence of abrasive particles in the lubricating oil on the surface of the friction pair. During the experimental process, with particle size and concentration in the lubricating oil, the rolling-sliding ratio, and the experimental duration as variables, contact fatigue tests are conducted through the coordinated operation of the JPM-1 type testing machine and the lubricating oil circulation system. After the test, KEYENCE VK-X250 series laser microscope is used to observe the surface of the specimen, and combined with 3D topography measurement, the two-dimensional and three-dimensional related parameters are extracted and analyzed. The results show that when the particle size of iron oxide particles increases, the surface roughness and protrusion parameters of the specimen decrease first and then increase, and the volume and number of pits continue to increase. The increase of concentration aggravates the wear, and the roughness parameters, the volume and surface area of the pit convex increase. At the same time, the increase of the slip-roll ratio and the extension of the test time will aggravate the wear, resulting in a significant increase in the coefficient of friction and the amount of wear. This experimental study can provide a reference for the mechanism of surface damage caused by abrasive particles in line contact parts.

Knowledge graphs (KGs) provide rich semantic and structured knowledge to address data sparsity in traditional recommendation systems, yet efficiently leveraging KGs for better performance remains challenging. This paper proposes KIDRec, a knowledge graph recommendation framework integrating knowledge interaction and diffusion-augmented learning. This framework captures deep KG semantics via high-order entity interactions and uses a diffusion model to generate robust knowledge representations through “noise injection and denoising” on collaborative user embeddings and item embeddings. It then incorporates an adaptive contrastive learning module with hard negative sample enhancement and a dynamic temperature coefficient. Experiments on Book-Crossing, Last.FM, and MovieLens20m show KIDRec outperforms baselines in AUC (0.782, 0.877, 0.979) and F1 (0.682, 0.796, 0.932), with relative AUC improvements of 0.031, 0.032, and 0.002, respectively, while alleviating data sparsity.

The role of attention in three distinct forms of numerical processing (i.e., subitizing, estimation, and counting) has been extensively studied. However, the similarities and differences in the impacts of top-down and bottom-up attention on these three processes remain poorly understood. This gap raises key theoretical questions: Do individuals adopt a uniform cognitive strategy (i.e., a static strategy) across forms of numerical processing and types of attentional modulation? Do they dynamically adjust accuracy and/or precision for varying forms of numerical processing and/or different types of attentional modulation (i.e., a dynamic strategy)? Or do they exhibit greater flexibility by combining these two strategies, depending on specific effects of attentional modulation on numerical processing? Using a novel paradigm that incorporates counting with continuous attentional consumption, we identified a combination of static and dynamic strategies: A greater reliance on attention for processing precision of small numerosities is ubiquitous across numerical processing forms and attentional modulation types. However, an attention-driven transition effect occurs exclusively across forms of numerical processing, not types of attentional modulation. Additionally, attention modulation on central tendency effect differs across numerical processing forms and attentional modulation types. These results highlight the dynamic nature and flexibility of attentional modulation on numerical processing.

In the era of the rapid development of the Internet, the traditional teaching mode of Java programming course can no longer meet the needs of educational development in the new era. First, this article combines outcome-oriented education with AI technology to optimize the curriculum system and teaching resources, deeply integrating theory with practice to enhance students' comprehensive analysis of knowledge and skills and their ability to solve complex problems. Then, collaborative filtering algorithms is utilized to fully tap into students' development potential and provide them with a personalized learning process taking the eight main steps of "preparation, introduction, analysis, teaching, implementation, investigation, conclusion, and expansion" as the main thread. Finally, the assessment methods are comprehensively reformed to build a personalized evaluation system through ai generation technology, fully covering the management of teaching and learning. Practice has proven that this new form of ternary integration teaching, involving teachers, students, and machines, has certain practical value for the reform and innovation of new engineering education.

This study proposes a real-time automobile braking performance analysis framework based on big data and machine learning technology, aiming to improve driving safety and the intelligent level of vehicle performance evaluation. By designing an efficient data preprocessing process, including sliding average filtering and linear interpolation methods, combined with the long short-term memory network (LSTM) model, accurate prediction of braking distance and response time is achieved, with a mean square error as low as 0.25, which is better than the traditional model. At the same time, the system architecture adopts a distributed computing framework to ensure millisecond-level real-time reasoning capabilities. Experimental results show that a reasonable setting of the discard rate can effectively improve the generalization ability of the model, verify the applicability of the framework in complex driving scenarios, and provide technical support for the development of intelligent driving systems.