Medium-depth coaxial geothermal wells, as an efficient way to utilize geothermal energy, are now being vigorously promoted by many regions in China. In order to study its heat extraction performance and its influencing factors, COMSOL Multiphysics was used for modeling and orthogonal experimental analysis. The results show that the effects of all influencing factor on the heat extraction from geothermal wells are listed in the following order: ground temperature gradient, inlet flow rate, thermal conductivity of the backfill material, inlet temperature, and thermal conductivity of the inner tube. When designing geothermal wells, it is necessary to consider multiple factors to determine the optimal operating conditions. The radial temperature influence range of geothermal wells is about 10 m, which is a reference for the arrangement of multiple wells in the same area.

Serving as a clean and renewable energy source, wind energy plays a significant role in mitigating the increasingly severe energy crisis. However, the fluctuation and randomness of wind speed pose severe challenges to the stable operation of power systems. To address this issue, a combined short-term wind speed forecasting model named CEEMDAN-RIME-CNN-BiLSTM-AM was proposed, which was based on complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN), rime optimization algorithm (RIME), convolutional neural network (CNN), bidirectional long short-term memory network (BiLSTM), and attention mechanism (AM). Initially, the CEEMDAN algorithm was applied to the original wind speed series to obtain a series of relatively stable sub-modes, thereby reducing the volatility of the wind speed series. Subsequently, the CNN hyperparameters were optimized using the RIME algorithm to establish the CNN-RIME model for adaptive extraction and mining of wind speed data. Then, the BiLSTM-AM model was employed to forecast the processed data. Finally, the forecasting results of each sub-series were superimposed to obtain the final forecasting result. A comparative experiment was conducted using an actual wind speed dataset from a certain location. The proposed model demonstrates good forecasting performance in both single-step and multi-step forecasting, providing a reference for scheduling plans to maximize energy utilization and power supply.

The dispatchable potential of charging stations represents the feasible solution space for optimizing their bidding strategies in the electricity market. However, the uncertainty in the access times of electric vehicles complicates the accurate assessment of this dispatchable potential. To address this issue, an uncertainty analysis method was proposed for evaluating the dispatchable potential of charging stations, taking into account the stochastic nature of electric vehicle charging times. Firstly, a generalized energy storage model for various types of electric vehicle clusters was established using Minkowski summation theory. Secondly, the impact of the randomness in electric vehicle arrival and departure times on the dispatchable potential of charging stations was analyzed. The discretized probability density functions of these times were mapped to the probability distributions of individual electric vehicle model parameters. By integrating these with the generalized energy storage model of electric vehicle clusters, the probabilistic characteristics of the dispatchable potential across various electric vehicle clusters were derived. Furthermore, the probability distribution characteristics of the dispatchable potential of charging stations were derived by aggregating the parameters of various electric vehicle clusters using convolution operations. Finally, simulations were conducted in MATLAB and compared with Monte Carlo simulations to validate the effectiveness of the proposed method.

In order to solve the fault line selection problem of single-phase high resistance grounding in resonant grounded distribution systems, a fault line selection scheme was proposed that utilized the disturbance characteristics of zero-sequence current before and after the neutral point parallel resistor was grounded. The zero-sequence fault model was established for the two conditions before and after the resistor paralleling with the arc suppression coil, and the zero-sequence current variation characteristics of sound and faulty lines were analyzed correspondingly. The amplitude of the zero-sequence current of any healthy line decreased after the neutral resistance was applied, while the current of the faulty line increased. Furthermore, a fault line selection criteria was constructed, which was used the amplitude disturbance characteristics of the zero-sequence currents. Simulations based on MATLAB verified the correctness and effectiveness of the proposed method. The results show that the proposed scheme is able to reliably detect single-phase ground fault with the resistance up to 5 kΩ.

The unmanned aerial vehicle (UAV) system, with its advantages of flexible deployment and line-of-sight propagation, has become an essential tool for assisting mobile communications in handling high-density data processing and emergency communications. However, the computational processing capabilities and endurance issues of UAVs under complex environments remain significant technological bottlenecks. The development of mobile edge computing (MEC) technology provides an effective solution to address UAVs’ computational and energy consumption challenges. A distributed task offloading strategy based on a multi-agent reinforcement learning algorithm was proposed for MEC-assisted UAV systems. The task offloading and resource allocation process of UAVs was modelled as a Markov game process (MGP) involving multiple MEC nodes. To solve the MGP problem, a distributed reinforcement learning algorithm for multi-agent collaboration was proposed. The algorithm enabled agents to find the optimal strategies through online collaborative learning based on local observation information. In comparative experiments, the convergence and system performance of the proposed scheme were evaluated. The results show that the proposed scheme outperforms the comparison schemes in terms of convergence speed, energy consumption, and unloading rate.

Shield tunnel construction impacts old buildings in urban areas. Extensive building deformation data was collected for Tianjin Metro Line 7’s shield tunnel passing beneath old buildings using automated measurement robots. Machine learning algorithms were applied to analyze the correlation between building instantaneous settlement and shield construction parameters such as average velocity, thrust, grouting volume, shield distance, and grouting pressure. A predictive model for building settlement was established. The results show that old buildings within -50~70 m are affected by shield construction. Differential settlement is significant, with noticeable differences on various facades. Grouting volume, thrust, and average velocity positively correlate with instantaneous settlement, with shield distance having the greatest impact. Reasonable construction parameters ensure building settlement and deformation remain within acceptable limits. The machine learning-based predictive model closely aligns with actual settlement curves, demonstrating robust predictive capabilities. This provides valuable insights for predicting and controlling surface settlement in future shield tunnel projects.

In order to solve the problems of missed detection and false detection in the current remote sensing image small target detection task, a SMCA+CSC+shape-aware intersection over union loss(SIoU)-you only look once(SCS-YOLO) remote sensing image small target detection algorithm was proposed. Firstly, in response to the problem of small and clustered targets in remote sensing images, a spatial multi-scale convolutional attention module(SMCA) was constructed to improve the model’s feature extraction ability of spatial and channel information. Secondly, in order to solve the problem that the semantic information of small targets was easy to be lost during deep network transmission, the aggregation subpixel convolution module concentrated sub-pixel convolution(CSC) was designed, and the multi-scale aggregation feature extraction method was used to enhance the ability of the network to extract semantic information. Finally, the SIoU loss function was used to replace the complete intersection over union loss(CIoU) loss function in the original model, which accelerated the convergence speed of the network. The average of the average precision(mAP)of the SCS-YOLO model reaches 97% and 90.9% on the RSOD and NWPU VHR-10 datasets, respectively, which is 2.2% and 2.7% higher than that of the original model, which shows the effectiveness of the method in the small target detection task of remote sensing images.

Due to the limitations of traditional ice force measurement methods in terms of stability and reliability, and the high sensitivity and anti-interference capabilities of fiber optic sensing technology was given, a fiber optic ice force sensor was developed. The effectiveness of this sensor was evaluated in the context of its application in marine structures. Based on the fundamental principles of fiber optic sensing technology and the design requirements of the ice force sensor, the research, design, and installation processes of the sensor were described in detail, including the design calculations of the elastic element, the selection and arrangement of the fiber optic sensors, and the construction of the data acquisition system to ensure that the precision requirements for ice force measurement were met. A winter field measurement of ice force was conducted at an observation station in the northern Bohai Sea. Field ice force data were successfully collected and analyzed, and the actual monitoring performance of the sensor was evaluated. The experimental results indicate that the system exhibits good stability and reliability in practical applications. The developed fiber optic ice force sensor provides a new reliable technical means for ice force measurement in marine engineering and lays a foundation for further research in structural health monitoring.

Modern air traffic management systems necessitate efficient and accurate identification and classification of hazard-related text data to ensure flight safety. Air traffic control hazard data encompasses information on potential factors, conditions, or events that may adversely impact aviation safety. Existing text classification methods face challenges due to the diversity of data categories and imbalances within classes. An enhanced ensemble model based on the Stacking framework, incorporating a dual-weighting mechanism was proposed for improved performance. A dual-protection strategy was implemented to categorize hazards and safety risks systematically. The methodology employed the term frequency-inverse document frequency(TF-IDF)algorithm to extract and vectorize features from preprocessed hazard texts. To address class imbalance, the synthetic minority over-sampling technique(SMOTE) and adaptive synthetic sampling approach(ADASYN)algorithms were utilized to generate synthetic samples for minority classes. The Stacking ensemble model was refined by dynamically weighting the F₁ scores derived from cross-validation of base learners and integrating a sensitivity assessment mechanism across the ensemble. Experimental results on the constructed dataset demonstrate that the ADASYN-enhanced ensemble model achieves notable improvements in precision, recall, and F₁ scores by 0.9%, 1.1%, and 1.0%, respectively, effectively mitigating overfitting in majority classes. The proposed algorithm significantly enhances the classification performance of imbalanced hazard text categories, contributing to the advancement of safety risk management in air traffic control.

Due to the complex underground environment, low lighting conditions, and the small size of hard hats, the detection results are not ideal. To address low-quality images in complex environments, an improved YOLOv7 for hard hat detection in low-quality images from underground coal mines was proposed. Firstly, addressing the limitation that image features were susceptible to noise interference under low-light conditions, a multi-scale MELAN module was introduced. By constructing a multi-scale attention mechanism, broader contextual information was captured, thereby enhancing feature extraction and effectively suppressing noise interference. Secondly, the OD-SMP module was constructed using soft pooling and full-dimensional dynamic convolution in the backbone network, which reduced information diffusion in feature mappings, retained more contextual information, and enhanced the detection capability for small targets. Finally, to address the varying quality of detection samples caused by the complex backgrounds and environments with different lighting and distances in underground coal mines, Wise-IoU was used as the loss function. Experimental results show that the average precision of the improved model is 94.9%, which is 13.5% higher than the original YOLOv7 model, demonstrating better detection performance.