Binzhou Technician College

Purpose This study investigates how social media influencers (SMIs) impact mimicry consumption by applying the heuristic-systematic model (HSM). It explores how the roles of empathy expression and self-disclosure influence consumers’ consumption imitation. Design/methodology/approach A quantitative, cross-sectional survey design was employed. Data were collected from 475 Chinese social media users. PLS-SEM and PROCESS marco were used to test the hypothesized relationships. Mediation effects of SMI credibility and emotional attachment, and moderation effects of homophily were also tested. Findings Empathy expression and self-disclosure significantly enhance both SMI credibility and emotional attachment. SMI credibility, in turn, positively predicts mimicry consumption, supporting the heuristic pathway. In contrast, emotional attachment does not directly influence mimicry behavior, and homophily does not moderate the proposed relationships. However, emotional attachment significantly mediates the effects of empathy expression and self-disclosure on mimicry consumption, functioning as an indirect relational mechanism rather than a direct driver of imitation. Research limitations/implications This study relies on self-reported survey data, which may be susceptible to common method bias and social desirability effects, and lacks behavioral or experimental validation. The model centers on HSM mechanisms without examining alternative mediators. In addition, SMI-level characteristics were not incorporated. Future research should adopt behavioral or longitudinal designs, integrate additional psychological mechanisms and consider both relational and structural SMI attributes to provide a more comprehensive account of SMI-driven consumer behavior. Practical implications SMIs should prioritize empathy expression and authentic self-disclosure to enhance credibility and stimulate mimicry consumption. Long-term collaborations with SMIs who demonstrate consistent expertise, transparency and genuine product use are more effective than relying solely on follower size. Originality/value This study is among the first to apply the HSM to SMI marketing, offering a novel dual-process explanation of mimicry consumption. Unlike prior research that treats emotional engagement and credibility as isolated factors, this study integrates them into a cohesive cognitive-emotional framework. It also challenges assumptions by showing that emotional attachment and homophily may not directly drive consumer mimicry.

ABSTRACT Organic dyes in industrial wastewater pose significant environmental and health risks. In this study, four lignin‐based hydrogels—Mg–Al–H, Al–H, Mg–Fe–H, and Fe–H—were synthesized by incorporating sodium lignosulfonate with metal ions to develop an eco‐friendly adsorbent for dye removal. Four types of lignin‐based hydrogels were systematically characterized using a particle size analyzer, Fourier‐transform infrared spectroscopy (FT‐IR), and scanning electron microscopy (SEM). Incorporation of magnesium and iron ions enhanced the porous structure, forming honeycomb‐like surfaces with abundant active sites. Among them, Mg–Fe–H showed the highest removal efficiency for methylene blue (MB), achieving 94.87% under optimal conditions (1 g/L dosage, 360 min, 298K, pH 10). Adsorption followed a pseudo‐second‐order kinetic model (R 2 = 0.974) and the Langmuir isotherm (R 2 = 0.971), indicating chemisorption and monolayer adsorption. The excellent adsorption performance was attributed to pore adsorption, electrostatic attraction, hydrogen bonding, and π–π interactions. These results demonstrate the potential of metal ion–modified lignin hydrogels for efficient and sustainable dye removal from wastewater.

• A thermal-seepage coupling model for artificial liquid nitrogen freezing was proposed based on SFCC data. • Accurately simulates freezing under high seepage conditions. • The prediction error is kept below 5 % in both no-seepage and high-seepage cases. • Outperforms traditional EPM in accuracy and efficiency. • Enhances AGF reliability in high-permeability formations. Using SFCC measurements obtained via nuclear magnetic resonance, a temperature–pore ice content relationship was established and a numerical simulation method for a thermo-seepage coupling model that accounts for high-seepage effects was developed. The method’s performance was evaluated by comparing results to an analytical solution for a single-pipe liquid nitrogen freezing scenario without seepage and to experimental data from a three-pipe freezing model under high-seepage conditions. Additionally, the approach was compared to the conventional Enthalpy-Porosity Method (EPM). The results show that: under no-seepage conditions, the numerical model predicts temperature distributions and freezing front radii with deviations below 5 %; under high-seepage conditions, the temperature discrepancy between simulation and experiment remains within 2 °C. Compared to EPM, the proposed method significantly reduces errors in closure time and computation time across seepage velocities ranging from 2.5 to 15 m/d; moreover, at high flow rates, the freezing wall maintains a nearly horizontally symmetric closure, avoiding the offset and distortion observed with EPM. This method enhances the accuracy and physical consistency of artificial ground freezing (AGF) simulations under high-seepage conditions, while also improving numerical stability and computational efficiency, thereby providing a more reliable tool for engineering design and safety assessment of liquid nitrogen freezing in high-permeability formations.

Abstract The performance of catalytic microreactors is critically governed by external heat loss. This work presents a systematic numerical investigation into the effect of external heat loss on the performance of a catalytic microchannel reactor fed with a propane/air premixture. The findings reveal a strong dependence of the ignition characteristics on the convective heat transfer coefficient ( h s ) at the outer wall, which controls the rate of energetic accumulation within the microreactor. Intense heat loss delays ignition by suppressing pre-heating and reducing the maximum combustion temperature. Near-adiabatic conditions (low h s ) facilitate rapid ignition and high combustion efficiency, leading to high temperatures (∼2010 K). The analysis of the wall centerline temperature profile further identifies a more pronounced temperature gradient with an increase in h s . This work provides critical insights and design guidelines for optimizing thermal management in micro-combustion systems.

ABSTRACT Transition metal oxides (TMOs) such as MoO x , featuring a high work function and a wide optical bandgap, are competitive alternatives to p‐type a‐Si:H or nano crystalline silicon in silicon heterojunction solar cells to form silicon compound heterojunction (SCH) solar cells. However, the thermal instability of MoO x during the curing process of screen‐printing restrains its massive production in industry. In this work, silver grids are printed on the MoO x side of MoO x SCH solar cells, and the influence of annealing atmosphere and temperature on the performance of the solar cells is investigated. After annealing in O 2 or air, the device performance is significantly degraded, while it remains almost unchanged after annealing in N 2 at 136°C. A conversion efficiency of 22.40% is achieved on the SCH solar cells with screen‐printed Ag grids when annealed at 136°C for 40 min in N 2 atmosphere, which is equivalent to that of the solar cells with thermally evaporated grids. To reveal the annealing effect, systematic research is conducted on changes in optoelectronic property, contact resistivity, and compositional distribution of MoO x , brought about by annealing in N 2 , O 2 , and air at different temperatures. Oxygen vacancies and conductivity both increase after annealing in N 2 , contributing to more efficient hole carrier collection through defect state‐assisted band‐to‐band transition. However, dipoles formed at the c‐Si/MoO x interface during N 2 annealing, proposed according to the calculation of differential charge density, might hinder the hole transportation from c‐Si to MoO x . A 1‐nm Al 2 O 3 layer inserted between a‐Si:H(i) and MoO x is found to be effective for mitigating the V oc drop after annealing in air. The approaches exhibit great potential for implementing screen‐printing on MoO x SCH solar cells, rendering industrial production of MoO x SCH solar cells feasible.

This paper proposes a trajectory tracking method based on swarm intelligence perception algorithm to address the problems of insufficient trajectory tracking accuracy and limited anti-interference ability of grasping robotic arms in dynamic environments. This method constructs an integrated framework of “perception planning control”, which integrates the target pose data of visual sensors and the contact force information of force sensors through swarm intelligence algorithms to achieve real-time estimation of environmental dynamic information; Generate a smooth trajectory that satisfies the constraint conditions based on the estimation results, and design an adaptive controller to achieve precise trajectory tracking. The experimental results show that the tracking error of this method converges to within 0.3 mm in static scenes, the response time in dynamic scenes reaches 45 ms, and the error amplification in interference scenes is controlled at 12%. The comprehensive performance is significantly better than traditional PID control and robust control. The deep integration of swarm intelligence and multi-source perception has solved the problem of disconnection between perception and control in traditional methods, providing technical support for complex grasping tasks.

This study aims to develop a forecasting model that predicts the annual number of museum visitors by integrating structured museum-related data and unstructured sentiment data. While prior research has often relied on a single data type or traditional regression techniques, this study incorporates sentiment scores extracted from museum-related news articles and user comments to empirically assess the influence of external public opinion. Seven predictive algorithms including traditional models (Linear Regression and Random Forest Regressor) and deep learning models (RNN, GAN, CNN, LSTM, and Transformer) were evaluated for performance. Among these, the Transformer model demonstrated the highest predictive accuracy across all evaluation metrics (RMSE, MSLE, and MAPE) and was adopted as the final forecasting model. The results show that incorporating sentiment data significantly enhances forecasting precision, highlighting the substantial impact of media narratives and public sentiment on visitor behavior. This study offers a robust forecasting framework that integrates both structured and unstructured data, providing practical implications for sustainable museum planning and strategic decision-making.

In response to the problems of feature redundancy and insufficient dynamic adaptability in the traditional random forest algorithm for identifying the operating status of electromechanical equipment, this study proposes an improved method that integrates feature optimization and dynamic integration strategy. Implementing feature selection through the mutual information ReliefF hybrid algorithm, removing 37% redundant information while retaining key physical features such as vibration peaks and temperature gradients; Introducing a dynamic weight integration mechanism based on classification accuracy and margin, the decision tree weights can be dynamically adjusted by <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$30 \%-50 \%$</tex> when the sample is at an abnormal boundary; Build a parallel inference architecture for edge cloud collaboration, with a single sample processing delay controlled within 45 ms. Industrial test data shows that this method achieves a recognition accuracy of 96.3% under four types of working conditions, an improvement of 8.2% compared to traditional algorithms. The recall rate for minor abnormalities such as early bearing wear is increased by 12.5%, and the accuracy remains at 90.2% in a 20 dB noise environment. The research results provide a solution for predictive maintenance of electromechanical equipment that combines accuracy and real-time performance. In the future, the application scenarios can be further expanded through model lightweighting and transfer learning.