The low-altitude intelligent network (LAIN) serves as a fundamental driving force for the intelligent development of the low-altitude economy and aerial traffic systems. However, current low-altitude communication technologies still face significant challenges, including limited coordination capability and insufficient transmission reliability, which severely hinder their ability to support services in complex and weak networks. To address the above challenges, this paper proposes an intelligent integrated communication architecture for LAIN, which is vertically divided into three layers: the integrated network layer, the resource adaptation layer, and the low-altitude service layer. It incorporates key technologies such as cross-network collaborative transmission, flexible multipath scheduling, enhanced coding redundancy, and covert channel authentication. By integrating technology and architectural innovation, the framework aims to enhance communication efficiency and reliability. Experimental results demonstrate that the effective transmission rate of the proposed scheme is at least 1 times higher than that of single-link transmission, the decoding success rate of network coding improves by an average of 11.05% compared to traditional methods, and the transmission accuracy of the covert channel has been improved by at least 10% compared with other algorithms.

The water-cooled flow channel in the casing of air compressor motor for fuel cell was designed and analyzed in this paper. Firstly, motor-CAD was used for electromagnetic simulation to determine the heat sources and heating power. Based on the heating power, the coolant flow rate was determined, and the convective heat transfer coefficients of the air gap and the main contact areas of the motor were calculated. A fluid-solid coupling model of the spiral-flow-channel water-cooled motor was established in ANSYS-Fluent software. The temperature at the end-region winding was validated by measuring the internal temperature of the motor using a PT100 platinum resistance temperature detector. The temperature rises of the motor was evaluated for the original flow channel under normal temperature conditions (coolant temperature ≈75 ℃) and high-temperature limit conditions (coolant temperature ≈85 ℃). The results show that the original design could meet the normal motor operation under normal temperature conditions. Under high-temperature limit conditions, the original design caused local overheating in the motor's right end-region winding, exceeding the H-level insulation temperature rise limit of 125.00 K. An optimized flow channel design was proposed to reduce the rate of cross-sectional area increase, thereby slowing the decrease in fluid velocity. Compared with the original design, the optimized channel increased the flow velocity at the overheating region, enhancing the coolant-motor heat transfer. The simulations results indicate that under high-temperature limit conditions, the optimized channel had a smaller cross-section change rate and a slower velocity reduction. The inner wall temperature decreased from 356.50 K to 354.85 K, and the maximum end-winding hot spot temperature decreased from 433.09 K to 422.92 K. Compared with the original design, the optimized end-region winding exhibits a reduction of 10.17 K in the local hot spot temperature. Furthermore, the temperature rise is 124.92 K, which is below the specified limit of 125.00 K. This demonstrates a significant improvement in the cooling performance of the flow channel.

A closed-loop heat pump drying system based on air heat recuperation was designed to enhance drying performance. To optimize system efficiency, adjustable parameters during the drying process were investigated. Multiple experiments were conducted to analyze the effects of bypass ratio, drying chamber temperature, and circulating airflow velocity on the recuperation system, and performance comparisons were made with a non-recuperative system. The experimental results indicate that each of these three factors, including bypass ratio, drying chamber temperature, and circulating airflow velocity, has an optimal value. The recuperation system improved the specific moisture extraction rate (SMER) by 5.57%~37.40%, depending on the conditions. Compare with the non-recuperative system, the optimized recuperation system achieve an SMER enhancement of 7.71%~22.78%.

Aiming at the damage monitoring problem of typical marine composite material structures, a new method for monitoring the damage location of composite materials is proposed, and the relationship between static loading load and damage expansion coefficient is established. For damage location monitoring, based on signal spectrum and correlation coefficient, a Bayesian-based damage probability imaging method is proposed, and a multi frequency weighted damage imaging update method is proposed. For damage expansion research, three damage expansion coefficients are constructed according to the time-domain signal characteristics of ultrasonic guided waves and hyperbolic tangent function, and the damage expansion coefficients under different static loading loads are obtained. Experimental results show that the Bayesian-based damage probability imaging method can accurately locate the damage location; the variation law of damage expansion coefficient with increasing load provides technical support for the damage expansion of typical composite material structures.

To ensure the welding quality of hot-melt joints of high-density polyethylene (HDPE) pipe, this paper proposes a total focusing method algorithm based on the time reversal algorithm and instantaneous phase coherent weighting factor. The method first processes signals through the time reversal algorithm, then extracts the instantaneous phase from the signals to construct the instantaneous phase coherent weighting factor, and weights the amplitude of each pixel of the total focusing imaging. Imaging detection experiments are carried out on 1 mm, 2 mm, and 3 mm transverse through-hole defects in HDPE test block and 3 mm hole defects in HDPE pipe butt fusion joint specimens. The experimental results show that the proposed algorithm can effectively suppress background noise and artifacts, reduce the amplitude of near-field noise (0~10 mm depth) and artifacts by about 30 dB, improve the resolution of total focusing imaging. Compared with the TFM algorithm, the average signal-to-noise ratios of the four defects are improved by 9.6 dB, 11.8 dB, 4.9 dB, and 3.5 dB, respectively, and the average array performance indicators are reduced by 56%, 49%, 38%, and 48%, respectively. This provides an effective improved scheme for the ultrasonic phased array detection of defects in butt fusion joints of HDPE pipe.

To address the challenges of the single variable time headway (VTH) model in simultaneously meeting the requirements of constant-speed cruising and variable-speed car-following over long-distance, this study proposes a continuous and synthesized variable time headway (CSVTH) model based on an improved particle swarm optimization (PSO) algorithm. Firstlg, a CSVTH model is designed to integrate the advantages of three single VTH models for calculating safe time headway for car-following. By employing transition functions to enable smooth switching between different VTH models, the proposed approach achieves multi-scenario VTH computation and derives reasonable following distances. Second, the particle swarm optimization algorithm is enhanced through adaptive inertia weight and learning weight adjustments, which are then applied to optimize the slope parameters of the CSVTH model. Finally, a Simulink-Carsim co-simulation environment is constructed for car-following scenarios. Simulation results demonstrate that, compared to the non-optimized CSVTH model, the optimized CSVTH model reduces fuel consumption by 4.4%, improves ride comfort by 6.4%, and decreases following error by 5.3% across three car-following scenarios.

Accurate modeling between physical devices and digital twins in industrial digital twins requires synchronizing massive amounts of sensing data. This results in difficulty ensuring real-time virtual-physical mapping. To address this issue, this paper proposes a four-layer bidirectional closed-loop architecture suitable for digital twins, consisting of a device, federated, cloud, and application layers. We develop a distributed federated bucketized decision tree algorithm utilizing histogram-based gain for global aggregation based on this architecture. Additionally, we design a pruning algorithm with adaptive gradient-weight redistribution to accelerate model convergence. Experimental results on the dataset demonstrate that the proposed federated aggregation model achieves accuracy improvements of 10%, 11%, and 22% compared to baseline methods. It also achieves convergence improvements of 15%, 20%, and 27%. Moreover, the proposed model maintains additional communication overhead within 300 mJ, even under the worst channel conditions.

In highway construction projects, the uneven settlement of roadbed caused by the abrupt change of geological conditions in different soil transition zones has become increasingly prominent problem. To solve the problem, a highway construction project in the transition zone between rock and sandy soil was taken as the background, based on the finite difference method, a numerical model of layered filling was established, and the settlement rule and the stress concentration effect of rock and sandy soil foundation were analyzed. The research results show that the subgrade settlement of the rocky section is small, and the maximum settlement is 2 cm, the subgrade settlement of the sandy soil section is large, and the maximu m settlement reaches 15 cm. In the transition zone between rock and sandy soil, obvious stress concentration occurs due to the abrupt change of soil properties, and the maximum settlement gradient reaches 1.2 cm/m, which is far beyond the allowable value of the specification. The construction sequence and the side-end effect have a significant impact on the settlement and stress distribution of the upper subgrade. Based on the research results, the improvement suggestions for subgrade settlement in different soil transition zones are put forward, the measures including setting the gradient transition layer, optimizing filling sequence and adopting geogrid reinforcement, which can reduce the settlement difference to the allowable range of the specification (<5 cm). The research conclusion can provide theoretical basis and technical support for the design and construction of similar projects, effectively reducing the settlement difference of subgrade in different soil transition zone, and ensuring the stability of the subgrade and driving safety.

Accurate and reliable tunnel boring machine (TBM) penetration rate prediction has significant engineering value for improving construction efficiency and ensuring construction safety. Aiming at the limitations of the existing TBM penetration rate prediction model with poor accuracy and insufficient consideration of uncertainty during construction, an interpretable TBM penetration rate interval prediction method based on machine learning was proposed. Firstly, several sets of domestic TBM tunnel engineering data were collected, with rock uniaxial compressive strength (UCS), rock integrity coefficient (Kv), thrust force (TF) and cutter speed (RPM) were selected as the input features. An extreme gradient boosting (XGBoost) point prediction model was developed through the Newton-Raphson optimization (NRBO) algorithm and cross-validation strategy. The Shapley additive explanation (SHAP) framework was introduced to analyze the contribution of the feature parameters to the prediction results. Then, the uncertainty of the point prediction results was quantified based on the adaptive bandwidth kernel density estimation (ABKDE) method, and achieve the interval prediction of penetration rate. Finally, the validity of the model was verified by a case study of the Kerman water transfer tunnel project in Iran. The results of the study show that compared with the XGBoost model without NRBO algorithm, the prediction error mean square error (MSE), mean absolute error (MAE), and mean absolute percentage error (MAPE) of the NRBO-XGBoost model have been reduced by 13.9%, 19.1%, and 0.7%, respectively, and the coefficient of determination, R2, has been improved by 0.015 1; the feature importance rankings are UCS (0.415 6)>TF (0.155 4)>RPM (0.104 5)>Kv (0.004 7), revealing that rock strength is the dominant influencing factor of penetration rate; the proposed model outperforms the adaptive boosting (AdaBoost) and the random forest (RF) model in the interval prediction performance, the predicted interval coverage probabilities (PICP) of NRBO-XGBoost, AdaBoost and RF models, reach 92.1%, 88.4% and 90.2%, respectively, with better uncertainty quantification ability; in the engineering example validation, the predicted R2 reaches 0.967 6 and the predicted intervals cover the measured values, which confirms that the model has a good applicability to engineering.