In order to explore the relationship between the ratio of non-high-density lipoprotein cholesterol to high-density lipoprotein cholesterol (NHHR) and the risk of rapid decline in kidney function, data from the China health and retirement longitudinal study (CHARLS) conducted in 2011 and 2015 was utilized. It includes 4 055 participants aged 40 and above with a baseline estimated glomerular filtration rate (eGFR) of at least 60 mL/(min·1.73m²), calculated using serum creatinine and cystatin C levels. Rapid kidney function decline is defined as a decrease in eGFR of 3 mL/(min·1.73m²·a). A multivariable logistic regression model was used to investigate the association between NHHR and the risk of rapid kidney function decline. Additionally, restricted cubic spline and threshold effect analyses were used to evaluate the dose-response relationship. The results show that over the 4-year follow-up, 447 participants (11.02%) experienced rapid decline in kidney function. In the fully adjusted multivariable logistic regression model, those in the highest NHHR group (T3) faced a 1.94-fold increased risk of rapid kidney function decline compared to the lowest NHHR group (T1) (OR=1.94,95%CI:1.46~2.61). As a continuous variable, each unit increase in NHHR is associated with a 1.21-fold increase in risk (OR=1.21, 95%CI:1.08~1.36). The restricted cubic spline analysis demonstrates a near logarithmic curve saturation effect between NHHR and rapid kidney function decline (P_nonlinearity<0.05). It is concluded that with NHHR ≤3.52, the risk increases as NHHR rises, while NHHR >3.52 marks a saturation point. In conclusion, among middle-aged and elderly individuals, a higher NHHR is linked to a greater risk of rapid decline in kidney function, displaying a nonlinear relationship.

Machine learning technology is a hot research topic at present. It is widely used in various prediction, recognition and classification tasks with its strong learning ability and high versatility. The application of machine learning in computational structural mechanics was discussed, with emphasis on its role in material property prediction, structural damage analysis, improvement of traditional methods, constitutive equation establishment and differential equation solving. Through literature review, the advantages of machine learning algorithms such as neural networks, support vector machines and random forests in improving computational efficiency and design process optimization were summarized. It is pointed out that the combination of machine learning and classical computing methods provides a new way to solve engineering problems. Future research will focus on algorithm optimization, model improvement and interdisciplinary technology integration.

The efficiency, precision, and automation of fruit and vegetable picking are realized through the integration of multiple mechanical arms in multi-mechanical arm cooperative picking technology, effectively addressing the high costs and low efficiency associated with traditional manual picking methods. The research progress in multi-mechanical arm cooperative picking technology was summarized, and the framework of the multi-mechanical arm cooperative picking system was comprehended. In light of the decision-making challenges in cooperative picking, the cooperative methods and task allocation within cooperative picking task planning was analyzed, and the collision detection, obstacle avoidance strategies, and path planning techniques utilized in cooperative picking with multiple robotic arms was reviewed. The future development direction of multi-mechanical arm cooperative picking technology is outlined, with a proposed development trend that envisions the combination of machine and agronomy, human-machine collaboration, decision-making big models, and multi-algorithm fusion.

Recently, the vessel-shaped aquacultural farm has received much attention from the academia and industry because of its importance in promotion and sustainability of marine fishery. Polluted surroundings unsuitable for fish growth are always introduced by traditional multiple point mooring system. Therefore, single point mooring system is more suitable for the aquacultural farm. In recent years, mooring scheme optimization and performance assessment was frequently investigated in most existing literature, while studies on design of the single point mooring system are rare. Based on the requirement of reducing hull reconstruction, an external turret single point mooring system was proposed, which is composed of the turret device, the mooring anchor, the truss structure and the bearings. And their working principle were also clearly illustrated. Then, three dimensional potential theory was applied to obtain the hydrodynamic coefficients of the ship by use of the software AQWA. Finally, in order to verify the positioning ability of the mooring system under both operational and survival conditions, the direction of wind, wave and current was combined according to the rules, and time-domain coupling analysis of the farm-mooring system was carried out. From the numerical results, it can be seen that yaw, relatively large sway and roll motions will be induced due to different direction of the wind, current and wave. The mooring line tension is adequate and safe under all conditions, indicating that this proposed single point mooring system exhibits good positioning performance.

In order to address the issues of low accuracy and high missed detection rates in existing pavement crack detection algorithms, an improved pavement crack detection algorithm based on YOLOv8n, named YOLO-CD (YOLO-crack detection), has been proposed. The scale sequence feature fusion (SSFF) module and triple feature encoder (TFE) module from the ASF-YOLO architecture were utilized by the YOLO-CD algorithm to enhance the detection performance for multi-scale cracks and the perception capability of target features. Additionally, the coordinate attention(CA) mechanism was introduced at the end of the backbone network and in the neck network, with positional information embedded into channel attention, thereby strengthening the extraction capability of crack features. Furthermore, an additional P2 small object detection layer was added on top of the original three output layers of YOLOv8n, increasing the multi-scale receptive field of the network, allowing both global and local context information to be captured simultaneously, thereby improving the detection capability for small cracks in complex scenes. The original YOLOv8n detection head was replaced by the DyHead detection head, achieving the integration of scale, spatial, and task attention mechanisms, and further enhancing the network’s detection performance for cracks. Experimental results show that in the self-built PD-Dataset, the mAP50 of the improved YOLO-CD algorithm is increased by 4.1% compared to the original YOLOv8n algorithm. In the public dataset RDD2020, the mAP50 of the improved YOLO-CD algorithm is increased by 1.5% compared to the original YOLOv8n algorithm. Moreover, the algorithm’s detection speed is found to reach 89.9 frames/s, meeting the real-time requirements of pavement crack detection.

The accurate prediction of soil compaction parameters has practical significance for improving soil bearing capacity and reducing compressibility in geotechnical engineering. The existing models have certain limitations in prediction progress and engineering applicability, and ignore the quantification of model prediction uncertainty. Genetic programming (GP) was used to model and predict two important soil compaction parameters (optimal water content and maximum dry density) for 226 groups of soil compaction test data with extensive and representativeness. The optimal display models of optimal water content and maximum dry density were obtained respectively, and the prediction results were compared with the results of existing prediction models. The GP model was quantified by combining quantile regression method and uncertainty statistics. The results show that the compaction parameters are most affected by fine grain content and plastic limit, while the gravel content and liquid limit have the least influence on them. Therefore, in practical engineering, the optimal compaction effect can be achieved by preferentially adjusting the fine grain content and plastic limit, while the gravel content (CG) and the liquid limit have the least influence on them. Therefore, in practical engineering, the optimal compaction effect can be achieved by preferentially adjusting the fine grain content (CF) and the plastic limit in the soil. In addition, the quantile regression (QR) method provides 90 % confidence and the mean prediction interval (MPI) is less than 0.3.At the same time, most of the data fall within the range of uncertain bands, indicating that the GP algorithm has strong prediction ability and high prediction accuracy. This interpretable display model is more convenient for engineering applications.

As essential components in power conversion modules, rectifiers are extensively utilized in power supply systems such as inverters, where their operational reliability directly influences the overall system performance. In order to enhance the reliability of rectifiers, it is critical to conduct lifespan predictions for sensitive components, particularly rectifier diodes. A predictive model was proposed that employs an improved grey wolf optimization (GWO) algorithm to optimize the hyperparameters of a simple recurrent unit (SRU) network. Initially, a power cycling accelerated aging test was performed on the diode, followed by an analysis of its characteristic parameters, with forward voltage drop identified as the primary aging indicator. Subsequently, the improved GWO algorithm was applied to optimize SRU hyperparameters—such as learning rate, number of hidden layers, and iteration count—thereby establishing a hybrid predictive model. Finally, the model was trained and validated using aging test data, with predictive accuracy compared against alternative models. The results show that the proposed model achieves superior predictive accuracy, and the data-driven predictive approach enhances the precision of diode lifespan estimation compared to conventional analytical modeling methods, thereby contributing to enhanced operational reliability of rectifiers.

In order to improve the output performance of permanent magnet assisted synchronous reluctance motor (PMa-SynRM), a multi-objective optimization design method for external rotor PMa-SynRM based on kernel extreme learning machine (KELM) and fast non-dominated sorting genetic algorithm (NSGA-II) was proposed. Firstly, the preliminary design of the PMa-SynRM rotor magnetic barrier was carried out and the working principle of the PMa-SynRM was analyzed. Secondly, the influence of each design variable on the optimization goal was evaluated through comprehensive sensitivity analysis, and the main optimization parameters were selected. Thirdly, with high output torque, high efficiency and low torque ripple as the optimization goals, a surrogate model based on KELM was established. Finally, NSGA-II was used for global optimization, and the optimal solution was selected from the Pareto frontier generated by NSGA-II, which was verified by finite element analysis. The simulation results show that the average torque of the optimized motor is increased by 15.83%, the torque ripple is reduced by 60.27%, and the efficiency of the optimized motor is also improved compared with the initial motor, which verifies the effectiveness of the optimized design method proposed in this paper.

An adaptive predefined-time prescribed performance backstepping fault-tolerant control strategy is presented based on radial basis function (RBF) neural networks, event-trigger mechanism and hysteresis quantizer for the attitude control problem of quadrotor unmanned aerial vehicle (UAV) with actuator faults. Firstly, the dynamic model of the quadrotor UAV system was constructed, and the attitude model was reconstructed by incorporating the actuator fault model. Secondly, by designing a class of time-varying functions, the error variables required for backstepping control were transformed. Thirdly, the nonlinear function approximation capability of RBF neural networks was utilized to estimate derivatives of virtual control laws and the actuator fault with unknown parameters. Finally, to reduce the update frequency of the actuator, a combination of event-trigger mechanism and hysteresis quantizer was used to design the control input. Stability of the closed-loop system was demonstrated through Lyapunov stability theory. The effectiveness of the proposed algorithm was verified through MATLAB. It is concluded that the designed event-triggered quantized controllers have a lower update frequency compared to controllers designed using only event-triggered techniques.

Structural design is an important part of the architectural engineering design stage, which must ensure that the building is safe, reliable, economical, and durable. Artificial intelligence can replace structural designers with a lot of training and repetitive operations to find the optimal design results and improve design efficiency. In order to comprehensively understand the relevant research and application hotspots of artificial intelligence in structural design, the current research status of artificial intelligence in the three stages of scheme design, preliminary design and construction drawing design was summarized from the perspective of the entire structural design process. Through reviewing literatures, it is found that artificial intelligence methods such as expert systems, decision trees, annealing algorithms, genetic algorithms, neural networks, and linear regression have been widely used in the field of building structure design, which has brought new development directions and approaches. At present, artificial intelligence methods are more widely used in the design of aboveground structures, but less in underground structures (foundations, basements, etc.), and their application in underground structures needs to be strengthened. In addition, the quantitative translation technology of normative provisions is relatively mature, but the qualitative translation technology of normative provisions still needs to be broken, and it is necessary to strengthen the research on rule-based or machine learning-based natural language processing.