In order to determine the reasonable spacing of underground interchanges, a series of traffic simulation experiments with varying spacing cases was conducted in VISSIM using the Wenhui Street underground interchange of the Two Lakes Tunnel in Wuhan City as a practical case. The effects of interchange spacing, mainline and ramp traffic volumes and design speeds on crash risk at the diverging and merging areas of the underground interchanges were analyzed. Then, the crash risk variation at the diverging and merging areas of the underground interchanges was predicted using the four typical machine learning algorithms, i.e., extreme gradient boosting (XGBoost), support vector machine (SVM), random forest (RF), and multilayer-perceptron (MLP). Results show that when the spacing increases from 1.5 km to 2.5 km, travel time, average delay, average queue length, and traffic conflict rate decrease significantly, and the collision risk index of time-to-collision(TTC) increase significantly. When the spacing is above 2.5 km, the decreases in travel time, average delay, average queue length, and traffic conflict rate start to slow down, and so does the increase in TTC. When the spacing is 2.5 km and above, the overall traffic operation efficiency and safety of the underground interchange increase significantly. The XGBoost model can predict the crash risk variation reaching a precision of 88.3%. This study can be a theoretical support and practical case for the setting of underground interchange spacing.

Hydrogen, which produces only water during usage, is an excellent secondary energy source. However, its environmental impact should consider the primary energy sources used for hydrogen production, as well as transportation. The use of the grid can not absorb the abandoned photoelectrolysis water to produce green hydrogen and incorporate it into natural gas, and the use of natural gas pipeline network transportation can ensure the environmental protection and clean hydrogen energy. An optimal operation model considering the start-stop characteristics of proton exchange membrane (PEM) electrolytic cell was established. The model can obtain the optimal production plan when dealing with intermittent energy, hydrogen demand fluctuation and time-varying electricity price, and achieve the balance of time-varying electricity price, hydrogen production, photovoltaic output and operating cost. The production plan shows the load of electrolyzer in different periods, which verifies the correctness of the model. By changing the minimum load in the constraint condition, the results show that the proportion of standby and idle state decreases with the decrease of the minimum load, and the running cost also decreases slightly. When the critical value reaches 6.1%, the running cost no longer changes.

In order to effectively predict the fuel consumption of vehicles, improve fuel economy and promote energy saving and emission reduction, a Hyperband-CNN-BiLSTM-based motor vehicle fuel consumption prediction method was proposed. Firstly, based on the vehicle operating status data and fuel consumption data collected from the actual road test, the salient factors affecting the fuel consumption of vehicles were analyzed. Secondly, combining the powerful feature extraction capability of convolutional neural network(CNN) and the advantages of bidirectional long and short-term memory network (BiLSTM) in dealing with the time-series data, a combined model of vehicle fuel consumption prediction based on CNN-BiLSTM was constructed. Then, in order to improve the model prediction accuracy, the combined model was optimized by Hyperband optimization algorithm, and the vehicle fuel consumption influencing factors were taken as the model input features to train the model to realize the modeling and prediction of vehicle fuel consumption. Finally, CNN, LSTM, BiLSTM, CNN-LSTM and CNN-BILSTM were selected as comparison models to evaluate the effect of Hyperband-CNN-BiLSTM prediction model. The results show that compared with other models, the Hyperband-CNN-BiLSTM model has the smallest mean absolute error (MAE) and root mean squared error (RMSE). They are 0.057 69 and 0.119 25, respectively. R² is the largest (0.991 76), and the model has the best prediction effect.

The issues of idler blocking during belt conveyor operation were addressed, which leads to excessive friction and abnormal temperature rise between idlers and conveyor belts. A friction surface temperature rise model for faulty idlers and conveyor belts was established based on microscopic friction theory, considering the phenomenon of hysteresis-induced heat generation and utilizing the virtual work approach. The finite element method was employed to conduct a thermo-mechanical coupling simulation on the friction model to analyze the effects of belt speed and load on temperature rise. An experimental platform was constructed to investigate the heat generation from friction between faulty idlers and conveyor belts, where an infrared thermal imager was utilized to monitor the temperature rise under varying parameters. The results indicate that the friction-induced heat generation between faulty idlers and conveyor belts positively correlates with both belt speed and load. An increase in either factor results in heightened heat generation, with the heat being primarily concentrated on the surface of the faulty idlers. The maximum deviation between experimental values and theoretical calculations is 8.7%, confirming the reliability of the theoretical model. Corresponding measures are proposed based on these findings.

Considering the high-temperature thermosetting chemical issues in the manufacturing process of composite tensile armor layers, the curing kinetics of T700/epoxy prepregs were explored. Through differential scanning calorimetry (DSC) analysis and the Starink method, the autocatalytic reaction curing kinetic parameters were accurately calculated. Then a curing kinetic model was established. It has been shown by experimental results that the reaction rate of the prepreg is significantly increased at higher heating rates. After the peak is reached, the reaction rate is decreased more rapidly, resulting in a lower average final reaction heat. The apparent activation energy of the curing reaction for this prepreg is 77.04 kJ/mol, and high consistency with the experimental data is exhibited by the constructed curing kinetic model.

China’s tight oil reservoirs have distinctive characteristics, including thin interbedded layers with alternate distribution in the longitudinal direction and strong reservoir heterogeneity. In order to maximize productivity and economic benefits, a development approach was commonly employed, involving a well network with layered fracturing for the simultaneous development of multiple layers. However, existing productivity models for fractured directional wells are only applicable to single-layer development and do not consider inter-layer interference, making them unsuitable for predicting well productivity of multi-layer development. In order to improve the accuracy of productivity prediction, the flow field nearby the fractured directional well is divided into the main fracture region, the stimulated reservoir volume region, and the un-stimulated reservoir volume region. Considering the effects of flow patterns in different regions and stress sensitivity, and introducing a disturbance coefficient, a non-steady-state productivity prediction model for multi-layer fractured directional well in tight oil reservoirs was established. After validating the model accuracy, the influence of fracture half-length, fracture conductivity, threshold pressure gradient, stress sensitivity and reservoir heterogeneity on the productivity of fractured directional well was further investigated. The results indicate that the threshold pressure gradient, stress sensitivity and longitudinal heterogeneity significantly affect the productivity of fractured directional well. The larger the threshold pressure gradient, and the more significant the stress sensitivity and longitudinal heterogeneity, the lower the productivity of fractured directional wells. With the gradual increase in fracture half-length, fracture conductivity, and matrix permeability, the productivity of fractured directional wells increases, but each factor has its optimal range. The ranking of factors affecting productivity is as follows: matrix permeability, fracture conductivity, fracture half-length, threshold pressure gradient, longitudinal heterogeneity, stress sensitivity.

The format and content of items such as product names and specifications in the detailed section of VAT invoices are highly flexible and complex, lacking complete gridlines to separate information fields. Existing methods for all-element structural recognition of VAT invoices face issues like low element recognition rates and high computational complexity. A structured recognition method for full face information based on computer morphology was proposed, which uses morphological operations to detect invoice table lines, cuts and recognizes text in different areas of the invoice. Then the implicit rules of the layout of the value-added tax invoice product details area was reused, combined with the text connected areas obtained through computer morphology operations, to construct a complete table structure. Finally, text detection and recognition were achieved using text detection neural network with differentiable binarization (DBNet) and convolutional recurrent neural networks (CRNN). The proposed method was tested on a dataset of 49 value-added tax invoices in three different formats, and the results show that the element recognition rates reached 99.9%, 97.4%, and 98.8%, respectively. The average running time per invoice is 0.90, 0.47, and 0.82 s, respectively. The structural recognition performance of the entire invoice exceeded multiple comparison table recognition models and literature methods.

To reveal the complex relationship between the built environment and walking activity among older adults, a gradient boosting regression tree (GBRT) model was adopted, combined with multi-source data such as mobile signaling data, remote sensing image data, and point of interest (POI), to deeply explore the non-linear impact of the built environment on elderly walking activities and its threshold characteristics. The findings indicate that the built environment has a significant nonlinear impact on older adults’ walking activity, with land use factors being the most influential. Specifically, land use mix, the proportion of commercial service facilities, and the proportion of residential land are identified as key factors affecting older adults' walking activity. Additionally, the proximity of facilities also plays an important role. Finally, suggestions have been put forward for the adaptive transformation of land use and facilities to improve the level of elderly walking activities and promote healthy aging.

Proppant performance is very important to the hydraulic fracturing design of unconventional oil and gas reservoirs. Few scholars have studied the micro performance parameters of proppant in terms of particle size and shape. The effect of particle size and shape on proppant breakage rate and fracture conductivity in shale gas reservoir was quantitatively characterized through laboratory experiments. The results show that when the closing pressure is lower than 28 MPa, the same type of proppant with uniform particle size and high spherical degree is compared with the proppant with poor sorting, the crushing rate is reduced by 15% and the fracture conductivity is increased by 10%. When the closing pressure exceeds the compressive strength of the proppant, the well-separated proppant can maintain the fracture conductivity better as the flow channel is further blocked by the debris generated by the broken proppant. The experimental results provide a reference for in-situ fracturing design of shale formation, improving the quality control level of downhole materials and selecting proppant.

In order to enhance the overall performance of the heat sink, a novel meshed microchannel heat sink structure was introduced, and its geometric parameters were optimized by performing a multi-objective optimization. The Box-Behnken design method was utilized to conduct response surface analysis on the design variables of channel width, fin thickness, and channel depth. The resulting temperature and pressure drop functions of the spider-shaped microchannel were then fitted as objective functions. The Pareto solution set was derived by applying a multi-objective particle swarm optimization algorithm, followed by utilizing the technique for order preference by similarity to an ideal solution(TOPSIS) method for selection from the Pareto solution set. It is concluded that the Pareto solution set is the optimal choice across various conditions. The multivariate statistical coefficients R² for temperature and pressure drop functions are 0.999 6 and 0.998 4, respectively, suggesting a high level of accuracy in the fitting function. The optimized structure not only reduces the average temperature by 3 K compared with the original design, but also decreases the pressure drop by 1 514 Pa. This significant improvement in comprehensive performance demonstrates that a well-designed channel structure can further enhance the heat sink performance of the microchannel.