This paper firstly starts from the thermodynamic theory, based on the classical heat transfer theory, and adopts the finite difference dichotomy method for mathematical modeling, and uses the secondorder center difference format to discretize the space, and solves the non-Fourier heat conduction equation. After completing the algorithmic solution of thermodynamic theory and finite difference method, the two are combined to deeply analyze and discuss the thermodynamic behavior of highspeed mechanical devices represented by high-speed rotating bearings. When the bearings operate at high speed, with the increase of stiffness, the pressure change in the middle and rear part of the bearings gradually flattens out, the temperature gradually rises, and the relative bearing capacity of the bearings decreases. The increase in the number of bearings also brings about an increase in the pressure at the centerline of the bearings, and the temperature of the air film corresponds to the increase in the average pressure, and there is a risk of over-temperature. In the thermodynamic characterization, the work done by the air film under compression and the heat generation due to viscous shear will lead to an increase in the temperature of the air film, which will lead to the temperature rise of the bearings, and will have a very great impact on the bearing performance.

Water resource is a high degree of unity between quantity and quality, once the water body suffers from pollution will make the water resources more scarce, and karst groundwater resource is one of the main water resources in the seismic area. In this paper, we chose Baiquanquan area in the low-mountain hilly area at the eastern foot of the south section of Taihang Mountains in H province as the research object, set 25 sampling points and collected 20 groups of karst groundwater samples and 5 groups of surface water samples, and carried out the reliability test by the ion balance method to control the error within ±5%. Based on the karst groundwater samples, the general characteristics of its hydrochemistry were analysed, and its hydrochemical characteristics were explored by cluster analysis. The causes of hydrochemical ions in karst groundwater were investigated by Gibbs plot, chlor-alkali index and saturation index, and the related factors affecting the hydrochemical characteristics of karst groundwater were investigated by factor analysis. The hydrochemical cations and anions in karst groundwater were mainly composed of Ca²⁺ and HCO₃^–, and the average concentrations of the two were 132.15 mg/L and 193.66 mg/L, respectively. The cast points of karst groundwater all fell between the dolomite and calcite areas, and their Mg²⁺/Ca²⁺ values ranged from 0.11 to 0.75, and the contribution of the F1 factor composed of Ca²⁺, Mg²⁺, SO₄^2-, TDS, HCO₃^– was the maximum of 38.91%. Karst groundwater in the seismic area will be affected by rock weathering, human activities, etc., which will affect the flow path of karst groundwater, and then have an impact on the hydrochemical composition of karst water.

In the context of building a new type of energy system, pumped storage projects have been widely adopted as a form of energy storage with the most mature technology and the most economical investment. In this paper, a hybrid pumped storage project online monitoring system based on multisensors is proposed, and an online monitoring database is designed and constructed. Based on the data in the online monitoring database, the soil erosion of the hybrid pumped storage project is calculated and analyzed by combining the CSLE model. Then the attention mechanism is combined with BiLSTM model to construct the landslide risk prediction model of hybrid pumped storage project. The soil erosion during the construction of the hybrid pumped storage project is mainly distributed on the construction land, garden land, grassland and cropland, among which the construction land has the largest area of soil erosion (132.19 km²), followed by the area of soil erosion of cropland (29.24 km²). The MAPE is between 0.002% and 0.005% when predicting landslide risk deformation of hybrid pumped storage project using CNN-BiLSTM-ATT model. And using the model in this paper can minimize the error of rainfall on the prediction of landslide risk deformation and realize the safe and stable construction and operation of hybrid pumped storage projects.

Based on the relevant theoretical foundation, a preliminary natural disaster public opinion risk evaluation index system is formulated, which consists of 4 first-level indicators, 15 second-level indicators, and 49 third-level indicators, and correlation analysis is utilized to screen the preliminary evaluation indexes, and after the screening of the indexes, the final system consists of 4 first-level indicators, 14 second-level indicators, and 23 third-level indicators. Indicator system. Taking a certain province as the research object, we synthesize the hierarchical analysis algorithm and fuzzy theory to explore the risk of natural disaster public opinion in a certain province. The natural disaster public opinion risk assessment result based on fuzzy theory is 69.48, while the corresponding affiliation matrix S_a = (0.2195, 0.3385, 0.1444, 0.1264, 0.1713), according to the principle of maximum affiliation degree, concludes that the natural disaster public opinion risk of the province is at the level of IV, and in order to improve the natural disaster public opinion risk management, the target strategy is proposed. This study has important reference value to promote the rationalization of natural disaster public opinion risk management, so that it can better prevent natural disaster public opinion risk.

In natural orchard environments, tangerines are susceptible to being shaded by foliage and to overlapping with multiple fruits. Varying weather conditions can cause inconsistent levels of illumination, and these unstable factors combined with complex backgrounds can diminish the efficiency of tangerine recognition and localization. Consequently, this paper utilizes images of tangerines captured under various weather conditions within a tangerine orchard as a dataset, and a method based on the YOLOv8n object detection algorithm is proposed. The dataset was trained using BiFPN, MCA attention mechanism, and PConv. An improvement in the algorithm resulted in an accuracy rate of 94.4% for tangerine target detection, a recall rate of 92.7%, an F1 score of 93.5%, and a mAP of 98.3%, with each metric showing an increase of 0.7%, 0.6%, 0.7%, and 1.3% respectively over the original model.

As the throat of transportation system, bridge structure is a lifeline project related to the coordinated development of society and economy. Based on fuzzy mathematical theory, this paper adopts Gaussian subordinate degree function to quantitatively characterize the damage detection information of articulation joints, combines Latin hypercubic sampling method and response surface method, and proposes a reliability assessment method for bridges integrating transversely distributed damage information, and analyzes the bridge of a certain structure as an engineering case, compares the failure probability corresponding to the model outputs and the real damage degree under different damage degrees, and analyzes the change of reliability indexes with the damage degree under articulation joint damage and main plate damage conditions, respectively. Under the conditions of articulation joint damage and main plate damage, the changes of reliability indexes with the damage degree are analyzed. The results show that the narrower the width of the damage interval is, the closer the failure probability is to the value corresponding to the real damage degree, and the reliability of the bridge decreases with the increase of the articulation damage and the main plate damage, which illustrates the objectivity and reasonableness of the method proposed in this paper.

Intelligent power preservation automation terminal is an important part of the smart grid, which plays an important role in improving the power supply quality of the power system. In this study, the CAB module is utilized to extract the multimodal features of visible images and infrared images during distribution network line operation. The weighted feature maps of each part of the distribution network operation are obtained by fusing different modal features through the average fusion strategy to realize line quality monitoring and fault detection and localization. The batch normalization layer and Relu function are also used to improve the image feature quality extraction performance of the model, which is then piggybacked on the model to construct a digital platform for intelligent power protection. The empirical analysis of the case found that the power supply reliability rate of L power supply company increased from 85.060% to 99.87% after the application of the smart power protection digital platform, the average power supply restoration time of non-faulty sections in the grid can be shortened to 3.24 minutes, and the line loss rate in the distribution network has been reduced to a certain extent. This study carries out the exploration of the practical application of the intelligent power preservation digital platform system, which lays the foundation for the stable operation of the distribution network and the improvement of power supply reliability.

Distribution network line project acceptance is a key link in the quality control of distribution network line project, an important factor affecting the safe and stable operation of the distribution network, which directly determines the level of safe operation of the distribution network. In this paper, for the distribution network line manual acceptance time-consuming and laborious, rare quality defects found rate identification rate is low and other issues, to carry out visual positioning and image recognition based on the distribution network drone automated acceptance technology research. In order to optimize the spatial positioning, attitude sensing and target tracking of the UAV, five coordinate systems, including the world coordinate system, body coordinate system, and photocentric coordinate system, are selected for spatial transformation. Based on the visual localization of the UAV, the path planning algorithm for UAV distribution line inspection combined with the path acquisition scheme is proposed. Gaussian denoising and histogram equalization are performed on the UAV inspection collected images, and Sarsa reinforcement learning algorithm is applied to train the samples to improve the automatic identification capability of safety hazards and other security hazards in the distribution network inspection. Visualization and analysis of UAV distribution line inspection path. Combine the distribution network defects dataset for optimal training strategy selection for distribution networks. The automatic identification algorithm for distribution network defects proposed in this paper achieves a mAP value of 79.60% in the target detection experiment. And in multiple dynamic path planning, the UAV nodes are able to accomplish the path planning tasks in different environments.

With the construction and development of new power system, grid business presents high reliability, high security protection, high flexibility, massive access level characteristics, 5G as the frontier technology of wireless network access, with high speed, wide connectivity, low latency features and advantages, and diversified grid business communication needs are highly compatible. Based on the characteristics of 5G communication technology, this paper analyzes its practicality in the power system. The main two protocols of the current autonomous network routing protocol for power system are proposed, and the inter-cluster routing optimization of OLSR is carried out by using AO algorithm. Simulate the predation behavior of skyhawk, develop the search strategy in the optimization process of AO algorithm, and construct the mathematical model of AO optimization algorithm. A quasi-inverse solution is used on the basis of the inverse solution to further increase the population diversity and convergence speed of the AO algorithm, while an adaptive weight factor strategy is used to balance the global search and local exploration capabilities of the AO algorithm. Simulation experiments are utilized to investigate the performance of the IAO algorithm as well as the PDR and delay in the mobile scenario of the power system. Comparing the PDR of the three protocols at different expected delivery distances, IOLSR still maintains a delivery rate of about 28% at a distance of 350m-500m. The optimized IOLSR shows further reduction in delay compared to OLSR in most of the cases with an average delay of 10829.43ns.

Wide-area protection systems are capable of eliminating or mitigating the consequences of disturbances by obtaining multi-point information about the grid system through measurement and communication techniques, and power system control and protection systems. In this study, distribution data in the grid system is collected and preprocessed, the distribution state of the grid system is estimated using data fusion methods, and an optimization method for distribution state estimation based on distributed computing methods is proposed. Then the grid wide-area protection system is designed by combining the grid system fault diagnosis method. Simulation and example analysis results show that the grid wide-area protection system based on data fusion and distributed computing has good performance in processing grid data and detecting and localizing grid faults, and the maximum localization error of faulted line points is maintained within 0.770%. In addition, this grid wide-area protection system is able to accurately detect a certain circuit fault in a regional grid system where faults are frequent, avoiding large-scale power outages and ensuring the stable operation of the grid system. This study has important scientific significance and application value for grid multivariate data fusion modeling and real-time fault detection, and provides an effective widearea protection scheme for the grid system.