In order to improve the recognition accuracy of high-altitude nuts and reduce the false detection rate of bolts and nuts, a high-altitude nut recognition model based on improved YOLOv5 was proposed. Firstly, a new attention mechanism efficient multi-scale attention(EMA) was added to the backbone network to integrate more information. Secondly, in order to enhance the network’s feature extraction capability, bidirectional feature pyramid network(BiFPN) was used to replace the PANet of the neck network. Finally, structured intersection over union(SIoU) was used to replace the original loss function complete intersection-over-union(CIoU) to accelerate the convergence of the model and improve its classification accuracy. The results show that the improved model has better performance than the original YOLOv5 model. The accuracy of the improved model increases by 0.92%. The recall increases by 0.16%. The average precision 1 (mAP_0.5:0.5) increases by 0.53%. And the average precision 2 (mAP:0.95) increases by 2.26%. An actual recognition comparison experiment between the improved model and the original YOLOv5 model was carried out. The experimental results show that the improved model has better recognition performance, which reduces the missed detection rate and the false detection rate, and improves the actual recognition rate. The improved model can well meet the recognition and image data acquisition of high-altitude nuts. And it also provide a data foundation for subsequent nut maintenance.

A method for optimizing the control parameters of the sample point distribution state within the framework of the unscented transform (UT) for the unscented Kalman filter (UKF) was introduced. The issue of abnormal filtering performance arising from the state of sample point distributions was addressed by this method. A multi-strategy improved sparrow search algorithm(ISSA) was employed to finely tune the control parameters. The goal is to enhance the distribution of Sigma points, thereby improving the effectiveness of nonlinear approximations and ultimately enhancing the accuracy of filtering estimations. To address the shortcomings of traditional sparrow search algorithms, several refinements were implemented. Initially, a Cubic chaotic mapping was applied to diversify the initial population. Furthermore, during the exploration phase, a nonlinear adaptive convergence factor was introduced to balance the algorithm’s capacity for global exploration and local exploitation. Additionally, a wavelet mutation strategy was integrated into the follower phase to prevent blind adherence to specific paths and mitigate the risk of becoming trapped in local optima. Lastly, an adaptive t-distribution perturbation capability was introduced to strengthen the algorithm’s ability to perform wide-ranging global searches. The efficacy of the proposed ISSA was demonstrated through simulation experiments conducted on various test functions. The results consistently show that ISSA outperforms other methods in terms of convergence and solution accuracy. Furthermore, the benefits of ISSA are extended to the optimization of parameters within the UKF algorithm. Experimental outcomes indicate that the ISSA-UKF algorithm reduces the root mean square error (RMSE) of position by 52.2% and the RMSE of velocity by 21.9%, thus affirming the viability and effectiveness of the proposed enhancements.

At present, the performance of the traditional centralized coil antenna used for radio frequency identification(RFID) detection and localization of underground cables is insufficient, which seriously restricts the improvement of its detection and localization distance. A new type of high field strength distributed RFID coil antenna structure was proposed. Based on the derivation of antenna related electrical parameters, the magnetic field strength of the coil antenna was taken as the objective function, and its quality factor was fixed as the constraint condition. Particle swarm optimization algorithm was employed to optimize the number of turns of the coil antenna and the turn spacing between adjacent two turns. Finally, an experimental test platform was built. The test results show that compared with the traditional centralized RFID coil antenna, the distributed RFID coil antenna increases the reading distance by 33.3%, significantly enhances the received signal strength indicator (RSSI) at the same distance, and helps to improve the accuracy of the underground RFID localization method based on RSSI, which provides an important reference for the application of RFID detection and localization of underground cables.

The Loess Plateau, as a natural ecological barrier in the western region of China, has made positive contributions to the sustainable development of the nation. The governance and restoration of the ecological environment on the Loess Plateau (Gansu region) plays a critical role in the implementation of China’s ecological civilization construction strategy. To monitor the changes in forest resources on the Loess Plateau (Gansu region) from 2008 to 2018, based on cloud platform, Landsat, PALSAR, and terrain data were integrated to explore the advantages of spectral index, backscatter, texture, and terrain features in obtaining forest resource information. The random forest feature selection algorithm was utilized to obtain the spatiotemporal distribution of forest cover in the study area for 10 years, and factor detection was conducted using geographic detectors. The results indicate that the random forest feature selection algorithm can effectively screen important feature information, with an overall accuracy of 91.88% and a Kappa coefficient of 0.91. The experimental scheme that integrates Landsat, PALSAR, and terrain data presents significantly higher accuracy compared to the forest classification results using a single data source. The overall accuracy of the four classification results is 86.65%, 88.23%, 90.15%, and 89.86% respectively. Over the past 10 years, the net increase in forest area in the study area is 0.60×10⁴ km². The areas with increased forests are primarily distributed in the central and eastern parts of Qingyang City, Pingliang City, Tianshui City, and the western region of Linxia Hui Autonomous Prefecture, while forest degradation primarily occurs in the southwestern part of Dingxi City and the central and eastern areas of Linxia Hui Autonomous Prefecture. In single-factor detection, land use type is the dominant factor in forest cover change, and the spatial distribution of suitable soil type and the auxiliary effect of rainfall provide favorable natural conditions for the survival rate of afforestation and the healthy growth of forests.

In order to study the solution properties of polyacrylamide at high temperatures at the molecular-atomic scale, the molecular models of partially hydrolyzed polyacrylamide(HPAM) and AM/AANa/AMPSNa copolymer [P(AM/AANa/AMPSNa)] have been established through a combination of experimental methods and molecular dynamics simulations. The solution properties of the two polymers at elevated temperatures were systematically investigated in terms of polymer chain rigidity and flexibility, hydrogen bonding, hydration layer, interaction energy and the effect of salt cation, and the micro-mechanism of temperature resistance of P(AM/AANa/AMPSNa) was explained. The results show that the introduction of side chains containing methyl and sulfonated groups into the molecular chain of P(AM/AANa/AMPSNa) could increase the rigidity of the molecular chain, more hydrogen bonds are formed between sulfonated groups and water and has longer lifetime. At the same time, the strong polar sulfonic acid group makes the hydration layer denser, which results in the weaker static shielding effect of the cations on the P(AM/AANa/AMPSNa). Under different temperature conditions, P(AM/AANa/AMPSNa) has stronger intermolecular non-bonding interactions, stronger water retention effect of the molecular chain at the microscopic level, and higher viscosity at the macroscopic level.

Primary cracks and new cracks develop within the engineering rock mass, leading to the formation of macroscopic cracks. The hollow cylindrical discrete element simulation test enables the emulation of complex stress paths. In order to solve the problems existing in the simulation test of hollow cylindrical discrete element, such as numerous influencing factors and lengthy meso-parameter calibration, a method of mesoscale parameter calibration of hollow cylindrical sandstone discrete element based on machine learning algorithm was proposed. Through variations in input variables within the discrete element model, 210 sets of simulation data were obtained. A mesoscopic parameter calibration model based on random forest algorithm and extreme gradient boosting(XGBoost) algorithm was established, the prediction accuracy of the model was compared, the parameter sensitivity was analyzed, and the contribution of input parameters to the overall mechanical properties of the rock was quantified. Combined with the indoor triaxial test of hollow cylinder, the calibration results show that the XGBoost algorithm has the advantages of computing speed, and can quickly locate the range of discrete element mesoscopic parameters, which provides a new idea for the calibration of discrete element mesoscopic parameters of hollow cylinder, and has the value of engineering application.

In order to solve the influence of intermittent and fluctuating wind energy on the economy and reliability of the system, a day-ahead and real-time energy management strategy for islanded wind power hydrogen production system with energy storage battery was proposed. In the day-ahead energy management stage, the energy management strategy of alkaline electrolytic cell power classification was adopted, and the economic energy management model of isolated wind power hydrogen production system was established with the maximum daily profit of the system as the objective function. An improved grey wolf optimization algorithm was proposed to solve the system energy management model, optimize the system energy flow and improve the system economy. In the real-time energy management stage, in order to ensure that the system can maintain stable operation when unplanned power fluctuations occur, a real-time energy management strategy for island wind power hydrogen production system based on condition identification was designed, which can adjust the system operation status in time according to the real-time operating conditions of the system. The experimental results show that the day-ahead energy management based on the improved grey wolf optimization algorithm effectively increases the daily revenue of the system and improves the economy of the system. The real-time stage energy management based on condition identification can maintain the system power balance when the wind power fluctuates under the day-ahead plan, and realize the stable hydrogen production of the electrolytic cell throughout the day.

The drilling process requires real-time measurement of drill string vibration, which is crucial for drilling and downhole safety. However, the traditional power supply mode used for downhole vibration sensors has been found to increase drilling costs and reduce drilling efficiency. Therefore, sensors with self-powered capabilities are considered more suitable for practical conditions. A downhole arrayed self-powered deformable vibration sensor was proposed based on the principle of triboelectric nanogenerators. Experimental results demonstrate that the sensor synchronized measurements of amplitude and frequency. The vibration frequency is measured within a range of 0 Hz to 11 Hz, with a measurement error of less than ±4%. Additionally, the sensor is able to measure three discrete amplitude values (10, 25, 40 mm) with a measurement error of ±3 mm. The sensor can working normally within a temperature range of 0 ℃ to 85 ℃. Furthermore, the sensor has power generation capabilities, with experiments revealing a maximum output power of 8.3×10^-7 W. Notably, when multiple sensors are used in parallel, the power generation capacity is significantly enhanced. These research findings provide new insights for the development of downhole sensors and downhole generators.

In order to investigate the properties of seawater mixing alkali-activated materials, the development law of hydration reaction, mechanics and corrosion resistance of seawater alkali-activated materials based on multi-component composite cementitious materials was studied. The results show that seawater mixing has a certain inhibitory effect on the hydration reaction of alkali-activated slag, and the compressive strength of SLCM at different ages also shows a decreasing trend to a certain extent, which is not conducive to the development of strength. Fly ash and silica fume can reduce the hydration reaction rate, early strength and toughness of the seawater alkali-activated materials, but their strength and toughness increase potential is significant in the later period, in which the strength and toughness growth rate of the silica-fly ash-slag terpolymer system from 7 days to 28 days reaches 50.9% and 86.7%, respectively. Compared with alkali-activated slag, adding fly ash and silica fume can improve the electric flux permeability and chloride ion mobility coefficient of alkali excited materials in seawater to a certain extent, which is consistent with their microstructure, but the three still belong to the same chloride ion permeability grade, i.e., medium permeability grade (electric flux method).

Aiming at the problem of effective extraction and identification of rolling bearing fault information in complex environments, a fault diagnosis method for rolling bearings based on feature mode decomposition (FMD) combined with multiscale fuzzy dispersion entropy (MFDE) and zebra optimization algorithm (ZOA) optimization support vector machine was proposed. In order to solve the problem that the key parameters in FMD are not adaptive, the minimum envelope entropy was used as the objective function, and the beluga whale optimization (BWO) was used to optimize FMD to find the optimal parameter combination to achieve the optimal decomposition of fault signals. Multiscale fuzzy dispersion entropy was introduced to construct the eigenvectors under different modes after decomposition. Finally, the feature vectors were input into the support vector machine for training and recognition. The effectiveness of the proposed method was verified by the public dataset and the self-made experimental platform dataset.