Bo Li

It is crucial to maximize the power output of wind farms to enhance the efficiency of utilizing wind energy. Turbine operational parameters have a significant impact on this. Nine Taguchi experiments are conducted, with each parameter having three levels: downstream spacing (3D, 5D, and 7D), yaw angle (10°, 20°, and 30°), and hub height difference (−0.2D, 0D, and 0.2D), where D is the turbine diameter. The total power output (Ptotal) under different combinations is measured. Range analysis and analysis of variance (ANOVA) are employed to rank the sensitivities of the parameters and identify the optimal configuration. Additionally, the velocity contour visualization helps reveal the underlying wake mechanisms that affect variations in power output. The results obtained show that downstream spacing (L) has the most significant influence on power output, accounting for 77.8% of the variance in the ANOVA. The yaw angle (β) follows, with a contribution of 16.0%, and then the hub height difference (Δh) with a contribution of 5.9%. Thus, the order of parameter importance is determined as L > β > Δh. The optimal parameter configuration is found to be L = 7D and β = 10°, which has been proven to result in a relatively high total power output. The velocity contour analysis verifies that this configuration reduces wake interference, optimizes redirection, and separates the wake from the downstream turbine's rotor plane, thereby enhancing performance. These findings provide valuable guidance for optimizing wind farm layout, emphasizing L as the primary control parameter, and suggest that adjustments to β and Δh can enhance the wake management and the energy capture efficiency.

Accurate prediction of wind turbine wakes is essential for optimizing wind farm layouts. This study proposes a novel analytical wake model (3DAESG) that integrates an elliptic super-Gaussian function with an adaptive parameter selection strategy using a genetic algorithm. Unlike conventional Gaussian-based models, the 3DAESG model captures the transition from a top-hat profile in the near-wake region to a Gaussian shape in the far-wake area. Validated against four distinct datasets—field measurements, wind tunnel tests, and large-eddy simulation data—the model achieves a maximum prediction error of less than 10%. Comparative evaluations demonstrate its improved predictive accuracy, especially in the near-wake zone and around the turbine hub height. Its computational efficiency makes it well-suited for practical wind farm planning and operation applications.

This study investigates the chain mediating effect of technology stress perception and career burnout on the relationship between generative AI tools usage and turnover intention among young tech talents. A cross-sectional survey design was employed with 683 technology professionals aged 21-35, using structural equation modeling to analyze the proposed sequential pathway. Generative AI tools usage significantly influences turnover intention through a sequential pathway: AI tools usage positively affects technology stress perception (β = .36, p < .001), which contributes to career burnout (β = .53, p < .001), ultimately increasing turnover intention (β = .49, p < .001). This chain mediation effect was significant (indirect effect = .094, 95% CI [.071, .124]), explaining 67.1% of the total effect. The findings extend technostress theory to generative AI contexts and establish that technology self-efficacy and organizational support function as protective factors by mitigating technology stress and burnout, respectively. Practical implications: organizations implementing AI technologies should adopt strategic approaches focusing on reducing technology stress and preventing burnout to maintain workforce stability during technological transitions.

The total mileage of the road network in China has been growing rapidly during the last twenty years. With the development of deep learning, the automatic road distr ess detection method is more accurate and effective than manual detection. However, the road are classified into five grades according to the Chinese road standard and each grade has its own characteristics. A single model cannot effectively identify multi-grade roads with different materials and levels of road distress. This study proposes a YOLOv8-based road distress detection strategy adapted for multiple road grades. The improved URetinex-Net network is used to enhance the spatial features and scenario diversity of the road distress datasets. Compared to the base YOLOv8 model, the enhancements have led to a 12% increase in accuracy for cement roads, a 22.3% improvement in detection speed, a 5.5% increase in accuracy for ordinary asphalt roads, a 7.5% increase in recognition accuracy for highways, and a 9.3% improvement in detection speed, with significant effects. This study refines the classification of roads based on their grades and matches them with corresponding artificial intelligence training strategies, providing guidance for road inspection and maintenance.

Calcareous sands often display wide ring grain configurations, high intragranular porosity, a complex structure, and low grain hardness. These attributes typically do not meet the strength criteria necessary to sustain overlying infrastructure in civil engineering applications. This study investigates gel stabilization techniques, blending gel material with calcareous sand at concentrations ranging from 5% to 22%, followed by curing periods of 3 to 28 days to evaluate the load-bearing capacity. Subsequently, an unconfined compressive test is performed to determine the gel material content in stabilized specimens and investigate the influence of gel material types. The gel material-to-sand ratios employed are set at 5%, 10%, and 16% for Portland cement and 13%, 16%, and 22% for gypsum. After that, a triaxial consolidated undrained test is conducted to assess mechanical behavior, pore water pressure, and mechanical properties. The findings reveal increased dilation, stress–strain hardening, and softening post-yield, regardless of gel material type. Principal stress ratios, secant modulus, and cohesion show a positive correlation with maintenance duration and binder content, with implications for improved load-bearing capacity. The study also elucidates the qualitative relationship between secant modulus E₅₀ and confining pressure.

The deep learning method has been widely used in the engineering field. The availability of the training dataset is one of the most important limitations of the deep learning method. Accurate prediction of pavement performance plays a vital role in road preventive maintenance (PM) and decision-making. Pavement performance prediction based on deep learning has been widely used around the world for its accuracy, robustness, and automation. However, most of the countries in the world have not built their pavement performance historical database, which prevents preventive maintenance using the deep learning method. This study presents an innovative particle swarm optimization (PSO) algorithm-enhanced two-stage TrAdaBoost.R² transfer learning algorithm, which could significantly increase the pavement performance prediction database. The Long-Term Pavement Performance (LTPP) database is used as the source domain data, and one of the highways in China is chosen as the target domain to predict pavement performance. The results show that the proposed PSO-Two-stage TrAdaBoost.R² model has the highest accuracy compared with AdaBoost.R² model and traditional regression decision tree model. The validation case study shows significant consistency between the predicted International Roughness Index (IRI) and the whole-year measurement data with an R² of 0.7. This study demonstrates the great potential of the innovative instance-based transfer learning method in pavement performance prediction of a region’s lack of data. This study also contributes to other engineering fields that could greatly increase the universality of deep learning.

Deep mixing method (DMM) has been widely used in geotechnical and highway engineering to stabilize soft soil foundations. The emerging material, one-part geopolymer (OPG), is environmentally friendly, which is promising to be used as an alternative binder to replace ordinary portland cement (OPC) partially or even completely for stabilizing the soft clay. In this paper, the dynamic shear modulus and damping ratio of the OPG-stabilized soft clay are investigated through the dynamic triaxial test. X-ray computed tomography (X-ray CT) is adopted to analyze the evolution of microstructure, especially the porosity distribution in OPG-stabilized soil samples. Factors affecting the dynamic behaviors of the OPG-stabilized soil, including the mass ratio of fly ash (FA) to ground granulated blast furnace slag (GGBFS), curing stress, curing time, confining pressure, cyclic-stress frequency, and cyclic-stress ratio (CSR) are studied. Results show that among all the mixing proportions, the sample with FA/GGBFS mass ratio of 0.1 could get the highest dynamic shear modulus and lowest damping ratio. With the curing stress increasing and curing time elapsing, dynamic shear modulus increases and the damping ratio decreases significantly. Furthermore, the higher the confining pressure is, the higher the dynamic shear modulus achieves. However, the higher confining pressure, such as 300 kPa, might disturb the original structure of the OPG-stabilized soil sample and then induce the breakage of bonding between hydration gels and soil particles, further causing the substantial loss of transmission energy. Finally, the higher cyclic-stress frequency would induce the larger shear strain of the sample under the given CSR, resulting in the lower dynamic shear modulus and higher damping ratio. The outcome of this study could provide guidance for the application of OPG binder in the ground improvement.