Research on the factors affecting soil organic carbon density is of great significance for regulating climate change and sustainable agricultural development. Previous studies have mainly explored the relationship between various factors (e.g., climate, altitude, soil physicochemical properties, etc.) and the influence of soil organic carbon density, but less involved in the interaction relationship between factors. Typical soil profiles were collected in Anhui Province to estimate the soil organic carbon density (SOCD) in the 0~10 cm, 10~20 cm, 20~30 cm and 30~100 cm soil horizons. The structural equation model was used to analyze the effects of climate, elevation, vegetation, soil water content, human activities and other environmental factors on SOCD. The results are as follows. In the 0~30 cm soil layer, SOCD show a gradually decreasing trend, and the average SOCD in the 0~10 cm, 10~20 cm and 20~30 cm soil layers were 2.09, 1.63 and 1.10 kg/m², respectively. The average SOCD of 30~100 cm soil layer is 4.46 kg/m². The spatial distribution of SOCD in the province gradually increased from north to south. The SOCD of 0~10 cm and 10~20 cm soil layer is higher than 5.00 kg/m^-2, mainly distributed in the Jianghuai hilly downland and the Riverine Plain. The areas with SOCD higher than 3.00 kg/m² in the 20~30 cm soil layer were distributed in the South Anhui hilly region. The high SOCD values of 30~100 cm are mainly distributed in the South Anhui hilly region. In the structural equation model of 0~10 cm, 10~20 cm and 20~30 cm soil layer, land use has the largest positive influence on SOCD, and the influence coefficients are 0.22, 0.20 and 0.22, respectively. The average annual temperature has the largest negative influence on SOCD, and the influence coefficients are -0.04 and -0.03. Annual rainfall was the most significant in 30~100 cm soil layer, but land use and NDVI were not significantly affected (p>0.05). Topography affects SOCD through four paths: land use, NDVI, annual precipitation and average annual temperature. Human footprint affected SOCD through NDVI, and the effect on NDVI reached a very significant level (p<0.001). The structural equation model established in this study initially explained the relationship between different environmental factors, and provid a theoretical basis for SOCD regulation and agricultural sustainable development.

An improved PSPNet(pyramid scene parseing network) network was proposed to automatically identify fractures in electrical imaging logging images, which was difficult to extract fracture features and led to low segmentation accuracy and large calculation of network parameters. Firstly, the backbone network in PSPNet was replaced with the optimized MobileNetV3 network, which could significantly reduce the number of network parameters and the amount of computation. Secondly, the asymptotic feature pyramid network(AFPN) was introduced to increase the interaction of multi-scale information and enhance the recognition ability of small cracks. Then, multi-depthwise Conv head transposed attention(MDTA) was introduced to extract global features and improve the extraction ability of key information. Finally, the combination of Focal Loss and Dice Loss were used as a loss function to solve the problem of unbalanced proportion of data sets. The experimental results show that the improved PSPNet network has a good segmentation effect on the fracture in the electrical imaging logging. Compared with the PSPNet network, mIoU(mean intersection over union) improved by 3.17% and mPA(mean pixel accuracy) improved by 6.38%. In addition, the number of parameters, calculation amount and weight of the proposed algorithm are reduced by 94.3%, 95.7% and 93.8% respectively compared with the original model. At the same time, the crack identification system based on CIFLog is developed, which can meet the practical needs of the electrical imaging logging.

The selection of variables affecting fuel consumption in the existing studies usually has no clear criteria, and it is difficult to combine the research results with actual flight. The flight training data of a Cessna 172 was used to predict the fuel consumption during the airborne phase of general aviation trainer aircraft. Firstly, based on the authors’ flight experience as well as correlation analysis, the features that influence fuel flow rate were selected from the pilot’s operational perspective. Secondly, a regression tree model was used to predict fuel flow rate under different flight conditions, correlating the aircraft’s actual flight status with the predicted fuel flow rate, in order to facilitate subsequent research on specific fuel-saving strategies from the flight technique perspective. Finally, a random forest model optimized with hyperparameter tuning was used to predict the fuel flow rate. The experimental results show that the accuracy of the model used is better than that of the existing research results, with a mean absolute error of 0.286 gallon/h, a root mean squared error of 0.496 gallon/h, a residual sum of squares of 0.968 4, and a mean absolute percentage error of 4.00%.

Signal processing and deep learning are often combined to achieve better diagnostic results in the field of fault diagnosis. Based on this, the symplectic geometric mode decomposition was improved and the ResNeXt neural network was optimized, and then a gearbox fault diagnosis model was proposed based on the combination of optimized symplectic geometric mode decomposition and ResNeXt neural network was improved. Firstly, the collected vibration signals were filtered and reconstructed by optimized symplectic geometric mode decomposition to obtain the effective components. Then it was sent to the improved ResNeXt neural network for fault recognition and classification. The rolling bearing variable condition data from the University of Ottawa was used to verify the feasibility of the model. The gearbox data from drivetrain dynamics simula (DDS) was used for contrast experiment and anti-noise experiment, which verified the effectiveness of changes and the generalization of the model.

Controlled-source audio-frequency magnetotellurics (CSAMT) uses artificial sources, providing strong anti-interference capabilities. It is widely used in oil exploration, mineral surveys and other areas. Traditional 2D inversion technology is mature, and deep learning has recently made some research advancements in geophysical exploration. There is still a research gap in applying deep learning to CSAMT inversion. Therefore, developing a 2D inversion algorithm for CSAMT based on deep learning is highly significant for advancing the use of deep learning in electromagnetic exploration. The characteristics of deep learning components such as convolutional layers, pooling layers, fully connected layers, and the UNet network were introduced. An explanation was provided on how to construct the training dataset, the UNet network used in this study, and how to set various training parameters. The network was saved after training. When the inversion was needed, the net was loaded and the algorithm could predict the result. Several theoretical models were designed for inversion, and the experiment results verified the reliability and effectiveness of the algorithm. The time of the deep learning inversion and the tranditional inversion was recorded. Building training set needed much time, but the time of deep learning inverison was much less than the tranditional inversion. The deep learning inversion is more efficient than the traditional inversion.

A hybrid algorithm (IWOA-BP) combining the improved whale optimization algorithm (IWOA) and backpropagation neural network (BP) was proposed to offer theoretical support for the formulation of grain strategies in the agriculture sector and its related industries. By introducing an improved convergence factor, nonlinear inertia weight, and optimal neighborhood disturbance strategy into the modified whale optimization algorithm, the optimal solution of the algorithm was obtained. This solution was then utilized as the initial weights and thresholds of the BP neural network, thereby enhancing the convergence speed and accuracy of the IWOA-BP hybrid algorithm. Subsequently, a grain yield prediction model based on the improved whale optimization algorithm was established using data from China’s grain yield over 45 years and seven influencing factors including effective irrigation area, chemical fertilizer application, rural electricity consumption, total power of agricultural machinery, sowing area of grain crops, disaster-affected area, and per capita consumption expenditure in rural areas. Through extensive experiments on a test set, it was found that the IWOA-BP model consistently outperformed other prediction models such as long short-term memory (LSTM), extreme learning machine (ELM), BP neural network with whale optimization algorithm (WOA-BP), and BP neural network with particle swarm optimization (PSO-BP). Compared to the ELM model, the root mean square error (RMSE) and mean absolute percentage error (MAPE) of the IWOA-BP model were reduced by 77.12% and 88.18% respectively. When compared to the LSTM model, the RMSE and MAPE of the IWOA-BP model were reduced by 69.11% and 47.36% respectively. Furthermore, in comparison to the WOA-BP model, the mean absolute error (MAE), RMSE, and MAPE of the IWOA-BP model were reduced by 43.78%, 43.22% and 45.96% respectively. Additionally, when compared to the PSO-BP model, the MAE, RMSE, and MAPE of the IWOA-BP model were reduced by 89.67%, 90.61% and 90.82% respectively. Therefore, the proposed IWOA-BP prediction model can be effectively used to predict grain yield due to its higher coefficient of determination, smaller prediction error, and faster convergence speed. It has important technical reference value for agricultural departments and relevant policymakers.

As a key device for rocket launch, the erecting device’s load-bearing performance is crucial for the success. In response to the instability phenomenon of the vertical plate of a certain rocket erecting device under the loading condition at the moment of erection, in order to analyze the reasons for the instability of the vertical plate, a finite element simulation model of the erecting device was established considering the actual load situation. The compressive load on both sides of the vertical plate under the working condition of erection was extracted, and based on the theory of small deflection thin plate elastic stability, the local instability of the vertical plate was explained: the compressive load of the vertical plate at the moment of erection exceeded its critical instability load, manifested as the characteristic of lateral bending deformation instability. Based on the analysis results, local reinforcement measures for the vertical plate were proposed. Finite element analysis and erecting loading tests were conducted on the vertical device after reinforcement. The results indicate that the stress consistency at the corresponding measurement points of the left and right vertical plates is good, and the stress deviation between the simulation calculation and the test result is not more than 10%. The lateral bending deformation decreases from 5.3 mm before reinforcement to 1.6 mm after reinforcement, proving the effectiveness of structural reinforcement. The relevant conclusions provide a reference for the local stability analysis of large and complex structures.

Since 2019, the Jiubaoyan landslide has exhibited continuous and gradual deformation. On September 17, 2021, during the rainy season, the landslide was obviously deformed and slipped due to the continuous heavy rainfall. On the basis of traditional engineering geological exploration methods such as on-site investigation, drilling and displacement monitoring, the finite element simulation method Midas GTS was utilized to simulate and calculate the seepage and displacement field of the slope under different working conditions, the landslide formation mechanism was comprehensively analyzed. Furthermore, the Fast GPU Matrix computing of discrete element method (MatDEM) was introduced to forecast the trend of the landslide sliding evolution under rainstorm working conditions. The results indicate these as follows. ① The finite element numerical simulation results are consistent with the drilling results, revealing that the sliding zone of Jiubaoyan landslide is located at the interface between the quaternary landslide accumulation layer gravel soil (${Q}_{4}^{del}$) and the mudstone of the Jurassic Suining Formation (${J}_{2}^{sn}$); ② Both finite element numerical simulation and on-site investigation suggest that Jiubaoyan landslide is a multi-level shallow soil landslide mainly driven by push forces and secondarily propelled by traction forces. The sliding mass is divided into upper and lower parts, which are closely interconnected. The loose soil structure in the landslide area as well as the abundant water storage ahead and behind the slope serves as material source for the landslide formation and annual heavy rainfall during the rainy season acts as critical external triggers for landslides.③ The finite element numerical simulation results reveal that Jiubaoyan landslide remains relatively stable under the natural conditions but transitions to an unstable state under the rainfall conditions. It is possible to lead to large displacement landslide under the persistent extreme rainstorms. ④ The discrete element simulation results suggest that the slope is in an destabilized state under extreme rainfall conditions, with two parts of sliding mass are penetrated through two sliding zones. This further landslide instability may culminate in significant displacement landslide, resulting in considerable economic and human losses. ⑤ The research method combining finite element analysis with discrete element analysis not only corroborates on-site investigation, drilling, and monitoring, but also enables a quantitatively analysis of the formation mechanism and prediction of potential working conditions in the future. It is hoped that such a study can offer some useful references for studying similar multi-level landslide disasters in mountainous regions.

To investigate the deformation characteristics of tunnels under impact and blast loads, A combination of model testing and numerical simulation was proposed to analyze the damage behavior and patterns under various dynamic loading conditions. The Hailuogou Tunnel was studied, and a combination of model testing and numerical simulation was utilized. Initially, an impact test was performed on the scaled-down physical tunnel model. Subsequently, numerical analysis of the tunnel model was conducted and verified. Then, a comparison was made between the tunnel deformation results of the scaled model and the prototype model under impact load. Finally, the effect of blasting load on the deformation of the prototype tunnel was analyzed. The results indicated that the proposed method accurately reflected the actual impact load’s effect on tunnel deformation. Additionally, the numerical analysis results of the scaled tunnel model closely matched the test results. Moreover, the deformation of the top of the prototype tunnel under the impact load was approximately 10 times greater than that of the scaled tunnel model, and it aligned well with the deformation caused by a blasting load equivalent to 500 kg of TNT. The impact load effectively simulated damage to the tunnel vault. Increasing the depth of cover and reducing the impact load represented effective measures to mitigate significant tunnel damage. The challenges of on-site testing during surface blasting are surmounted by this study’s findings. Additionally, a cost-effective, safe, and dependable testing approach is furnished for analyzing the destructive behaviors and modes of tunnels under various dynamic loads. Furthermore, technical support is offered for the safe and economical design of tunnels with optimized blasting loads.

To address the issues of long stitching time due to numerous mismatched feature points and insufficient stitching accuracy when using all feature points directly in image stitching tasks, an optimized image stitching method combining a matching point increasing strategy with RANSAC(random sample consensus) was proposed. The method initially screened feature points to prevent numerous ineffective samples, thus improving computational efficiency. Then, a progressive sampling strategy was employed to incrementally increase matching points and repeatedly sample for precise results. Finally, the optimal model was obtained by utilizing a new loss function based on root mean square error to filter the results. The experimental results indicate that, without a noticeable increase in time consumption, the interior point rate of the algorithm in this paper is further enhanced, the mean and root mean square errors of feature points have decreased significantly, the accuracy of image stitching is improved, the misalignment phenomenon at the stitching seam is effectively improved, and the stitching errors in image stitching tasks are significantly reduced.