To address the challenges of high-dimensional features, large computational demand, and difficulty in improving the accuracy of secondary return water temperature prediction models for heat stations, a secondary return water temperature prediction model based on the xtreme gradient boosting-artifical neural network(XGBoost-ANN) was proposed. The feature screening layer uses XGBoost algorithm to calculate the importance scores of the original data features and determine the main features that affect the secondary backwater temperature, thus reducing the complexity of the model and improving the computational efficiency. Three layers of feedforward ANN were trained by Bayesian regularization algorithm as the secondary backwater temperature prediction layer, and the initial weights and thresholds of the ANN model were optimized by grey wolf optimizer (GWO) algorithm. The weights and thresholds of the ANN model were represented by grey wolf position vector. The fitness function was introduced to evaluate the performance of each set of weights and thresholds to help the model avoid falling into local optimality at the initial stage of training, so as to improve the performance and generalization ability of the model. Experimental results demonstrate that the constructed XGBoost-GWO-ANN secondary return water temperature prediction model achieved significant improvements. Compared to the model before feature filtering, the root mean squared error(RMSE) is reduced by 26.8%, the R² is increased by 11.3%, and the model inference time is shortened by 46.1%. Furthermore, the optimization of the initial ANN weights and thresholds using the GWO algorithm improve the RMSE by 20.0% and the R² by 3.4% compared to the unoptimized ANN model. These results indicate that the accuracy and generalization ability of the proposed prediction model are effectively enhanced.

Hydrothermal management technology is conducive to solving the problems of proton exchange membrane fuel cell(PEMFC), such as large heat dissipation demand, slow cold start, and short life. The dynamic response test of a high-power fuel cell was carried out, and the correlation between temperature and humidity and effective output voltage was verified by Pearson correlation coefficient. The influence of water management and thermal management on fuel cell was analyzed by literature review, and the current hydrothermal management and modeling methods were summarized. Water management methods mainly include reaction gas humidification, internal structure design, and drainage control, but it isn’t easy to achieve accurate online water content detection and closed-loop control. On the other hand, thermal management technology is relatively mature. The water cooling method of the traditional heat engine and the temperature control strategy is used to control the water pump and fan in the thermal management subsystem so that the temperature of the fuel cell and the temperature difference of inlet and outlet cooling water are kept in a reasonable range. However, there is a strong coupling between temperature and water distribution in the stack, so the single temperature variable and water variable study can not truly reflect the influence of temperature and water content on the performance of fuel cell. In the future, it is the key to improve the performance of fuel cell, effectively improve the parasitic power of appendage, and prolong its service life by using efficient hydrothermal coupling technology and considering the influence of temperature and water content comprehensively.

Studying the characteristics of airport traffic flow fluctuation range is fundamental for efficient traffic management and control. Mastering these characteristics plays a crucial role in maintaining the stability and effectiveness of overall airport operations. Considering the irreversibility of time and the cumulative impact of traffic congestion, which occurs when airport traffic exceeds facility capacity within certain time intervals, a method for constructing an adaptive crossing network was proposed. From the perspective of complex network topology, both the overall characteristics of the network and the centrality of nodes were analyzed. The integrated centrality of nodes was calculated using the independent weighting coefficient method, enabling the identification of key time nodes that are core hubs of strong fluctuations within the network. The results show that the adaptive crossing network, mapped based on the traffic data from Beijing Daxing International Airport, exhibits characteristics of complexity and order, featuring scale-free properties, assortativity, and a distinct community structure. The time period from 21:20 to 22:25 (nodes 257~269) ranks highly across various centrality measures, indicating a significant fluctuation impact range, and thus, these nodes are identified as core hub nodes within the network. The integrated centrality synthesizes various topological centrality features of the network, and through quantitative analysis, effectively characterizes the strong fluctuation nodes within the network. This method provides a theoretical basis and practical reference for the optimization of airport traffic flow management and the study of abnormal fluctuations, offering a new perspective for enhancing airport operational efficiency and safety.

Uplift piles, in accordance with their structural properties, effectively sustain the structural uplift loads and have emerged as an efficacious solution to address the anti-floating issue. The precise determination of the internal forces within uplift piles is crucial for comprehending their load-bearing characteristics. Nevertheless, the tensile capacity of concrete is relatively feeble. Once the load attains a specific magnitude, its elastic modulus will decline, rendering the traditional axial force calculation methods inapplicable. By leveraging the optical frequency domain reflectometry(OFDR) strain measurement technology and conducting indoor model tests of uplift piles, the strain distribution and evolution patterns of both steel bars and concrete during the pulling process were analyzed. The alterations in the elastic modulus of concrete throughout the tension-failure process were thereby obtained. A method for optimizing the axial force calculation, which exploits the relationship curve between the concrete strain and elastic modulus, was put forward. This enables the accurate acquisition of the axial force of the pile body and its subsequent application in practical engineering projects. The test results indicate that under the condition of small loads, the OFDR technology can identify the locations where concrete cracks emerge based on the strain curve of the pile body. In the event of pile body failure under large loads, the elastic modulus of concrete can be rectified using the relationship curve between strain and elastic modulus. Compared with traditional calculation methods, the relative error of the axial force throughout the entire process can be confined within 5%. The viability of this approach has been corroborated in actual engineering endeavors, and the optimized axial force calculation exhibits enhanced precision.

As a large transportation hub, airport terminals have complex structures, and the evacuation efficiency becomes extremely important when an emergency occurs. To improve the evacuation efficiency of the terminal building, an improved A* algorithm has been proposed for selecting the optimal evacuation path based on the actual distribution of personnel. Firstly, the flow of personnel in the terminal building was simulated, and data on the distribution of personnel was obtained. Then the time-varying distribution of personnel in each area was considered in the cost calculation of path selection. Finally, the A* algorithm was improved in terms of traversal methods, network weights, other factors, and congested paths were replanned to avoid congestion. The results indicate that considering the distribution of passengers in the terminal improves the evacuation paths at each node, which leads to shorter evacuation time compared to traditional A* algorithm. It also allows for the avoidance of congested paths. This study can provide theoretical and methodological support for the rapid evacuation of passengers in the terminal under emergencies.

In order to smooth out the output fluctuation of wind power generation system, a hybrid energy storage dual-layer fuzzy control strategy based on wind power prediction was constructed by adjusting the control strategy of hybrid energy storage system (HESS) to meet the fluctuation limit of grid connection. Firstly, the improved complete ensemble empirical mode decomposition with adaptive noise (ICEEMDAN) was used to decompose the original wind power data. Secondly, the improved Adam algorithm and Transformer model were combined to predict each component, and the prediction results are superimposed as the final prediction result. Finally, based on the predicted wind power fluctuation state and state of charge (SOC) of the hybrid energy storage system, the dual-layer fuzzy control strategy was adopted to adjust the hybrid energy storage system to ensure that the overcharge and overdischarge of the hybrid energy storage system are reduced under the premise of smooth grid connection of wind power. The results demonstrate that the proposed control strategy achieves lower fluctuation indices in wind power output, ensuring reliable grid connection. Moreover, it maintains the SOC of the HESS within an optimal range, leading to an overall enhancement in the system's comprehensive performance.

With the increase of air cargo volume, cargo plans are frequently interrupted due to disruptions in cargo demand, so rescheduling flight schedules is the core issue for air cargo recovery. An air cargo recovery model based on spatio-temporal network method was proposed with the goal of maximizing the profits of airlines under the disturbance of temporary increase in demand. Aircraft routes, cargo routes and flights were reorganized in the model and the initial flight plan was preserved as much as possible by adding penalty factors. In order to verify the effectiveness of the model, the model was solved using CPLEX solver. The proposed spatio-temporal network-based air cargo recovery model was compared with the model in reference. The results show that the proposed model has significant advantages in computational efficiency and finding optimal values, and the advantages become more apparent with the increase of the case size. The sensitivity of the model's solution results to the time window width and aircraft carrying capacity was analyzed. The results show that the narrower the time window, the slower the solution speed, while as the time window width increases, the solution speed accelerates and tends to stabilize. As the carrying capacity of the aircraft gradually increases, the solving speed of the model becomes faster and tends to be stable.

Container terminal yard as the hub of the operation area on both land and sea sides, the space allocation will directly affect the overall operation efficiency of the terminal. In order to reasonably allocate the space of U-shaped yard to improve the efficiency of container loading, two blocks sharing one intelligent guided vehicle(IGV) lane were formed into a group. From the perspective of reducing lane congestion, considering the constraints of high and low workloads of adjacent operation areas within the block groups, and with the goal of minimizing the total loading operation time and balancing the number of containers in the block groups, the first stage of the export container allocation model was constructed, and an immunity compensation mechanism and Metropolis criterion were introduced. And an improved adaptive genetic algorithm was designed by introducing immune compensation mechanism and Metropolis criterion. Based on the allocation results of the first stage, a priority descending balanced stockpiling strategy was established to ensure the continuous operation of the yard crane, and the second stage of the container allocation model was constructed, and a heuristic algorithm was designed to solve the problem. The design of large-scale case experiments shows that the algorithm is able to give effective export container stockpiling area delineation and container space allocation scheme, and the performance of the algorithm is verified by different scale comparison experiments, which shows that the algorithm has a faster convergence speed and is able to optimize 10% of the target value, which improves the overall operational efficiency of the U-shaped automated terminal.

To address the challenges of complex geology, dense urban spaces, high component production requirements, and difficult excavation parameter control in the intelligent construction of ultra-large diameter shield tunnels, a systematic study of key links in the shield construction industry chain was conducted based on the experience and digitalization needs of completed and ongoing projects. building information modeling(BIM), big data, IoT, and artificial intelligence technologies were applied to build a digital architecture and establish a full lifecycle coding system. A data platform was developed to promote the digitalization of design data, intelligent segment production, tunnel intelligent excavation, and lifecycle management. The research shows that the parametric drive greatly improves the efficiency of digital design, optimizes the precision of segment production and drilling, and enhances the ability to control the whole life cycle. It can be seen that digital means effectively solve the technical problems in construction, improve the ability of construction control, and provide a feasible digital solution for the construction of large-diameter shield tunnel.

The natural orifice intervention using continuum robots faces challenges such as tortuous and narrow intervention paths, as well as compressive forces exerted by soft tissues in the orifice. To address the issue in the delivery process where existing planning methods struggle to balance multiple control objectives, resulting in difficulty in reaching deeper positions, an autonomous planning scheme based on residual reinforcement learning was proposed. The method enables the autonomous delivery of continuum robots through natural orifices. A feedback deviation model between the delivery posture of the continuum robot and the spatial state of the natural orifice was established to control the posture target during the delivery process. Simultaneously, a Markov model of the overall motion process of the continuum robot was constructed to train the reinforcement learning algorithm. A residual strategy, generated by combining posture feedback control with reinforcement learning control, was used to output the optimal actions for the continuum robot's delivery process. Experiments conducted in a simulated bronchial orifice show that the proposed method converges over 60% faster than existing methods and can plan smooth, collision-free trajectories for the continuum robot's intervention through the orifice, outperforming existing methods in several key metrics.