Search

The Autonomous Underwater Glider (AUG) is a kind of prevailing underwater intelligent internet vehicle and occupies a dominant position in industrial applications, in which path planning is an essential problem. Due to the complexity and variability of the ocean, accurate environment modeling and flexible path planning algorithms are pivotal challenges. The traditional models mainly utilize mathematical functions, which are not complete and reliable. Most existing path planning algorithms depend on the environment and lack flexibility. To overcome these challenges, we propose a path planning system for underwater intelligent internet vehicles. It applies digital twins and sensor data to map the real ocean environment to a virtual digital space, which provides a comprehensive and reliable environment for path simulation. We design a value-based reinforcement learning path planning algorithm and explore the optimal network structure parameters. The path simulation is controlled by a closed-loop model integrated into the terminal vehicle through edge computing. The integration of state input enriches the learning of neural networks and helps to improve generalization and flexibility. The task-related reward function promotes the rapid convergence of the training. The experimental results prove that our reinforcement learning based path planning algorithm has great flexibility and can effectively adapt to a variety of different ocean conditions.

As the foundation of human survival, the development of agriculture has always played an important role in social development. The quality of agricultural development determines the speed of social progress. With the development of computer science, using computer technology to solve problems related to agricultural development has become an important research direction in current computer development. In recent years, remote sensing detection has received widespread attention, and the application of remote sensing detection technology in agriculture can provide great convenience for the development of agriculture. Benefit from the development of deep learning, remote sensing detection has made gigantic achievements. However, we have to face some challenges. First, deep learning depends on large scale of data with annotations, which expends inestimable human resources. In addition, as the depth of detection network increases, the amount of parameters explosively extends. In this work, we carry out the research based on the two problems. At the beginning, an active learning method considering classification and localization task is proposed. Our method can choose some few but valuable images. It uses 34.48% amount of training set, up to 97.4% baseline performance, and realize the compression of the scale of dataset, which reduces the trouble of manual labeling. Due to imaging and other factors, there exists many small objects in remote sensing images. So we add the mixed convolution, dilated convolution, and mosaic data augmentation modules into HRDNet network. Experiments on the agriculture dataset indicate that the improved algorithm can obtain about 2% higher than HRDNet. To reduce the number of parameters, we adjust the weighted sum ratio of importance scores dynamically. With the pruning ratio of 80%, the model volume has only 184 MB, degrading 70%. Model compression accelerates the detection speed for seven times on NVIDIA AGX Xavier, with a speed of 6 FPS. Our work will lay a foundation for remote sensing detection.

Abstract Background Oocyte maturation arrest results in female infertility and the genetic etiology of this phenotype remains largely unknown. Previous studies have proven that cyclins play a significant role in the cell cycle both in meiosis and mitosis. Cyclin B3 (CCNB3) is one of the members of the cyclin family and its function in human oocyte maturation is poorly understood. Methods 118 infertile patients were recruited and WES was performed for 68 independent females that experienced oocyte maturation arrest. Four mutations in CCNB3 were found and effects of these mutations were validated by Sanger sequencing and in vitro functional analyses. Results We found these mutations altered the location of cyclin B3 which affected the function of cyclin dependent kinase 1 (CDK1) and led to mouse oocyte arrested at germinal vesicle (GV) stage. And then, low CDK1 activity influenced the degradation of cadherin 1 (CDH1) and the accumulation of cell division cycle 20 (CDC20) which are two types of anaphase-promoting complex/cyclosome (APC/C) activators and act in different stages of the cell cycle. Finally, APC/C activity was downregulated due to insufficient CDC20 level and resulted in oocyte metaphase I (MI) arrest. Moreover, we also found that the addition of PP1 inhibitor Okadic acid and CDK1 inhibitor Roscovitine at corresponding stages during oocyte in vitro maturation (IVM) significantly improved the maturation rates in CCNB3 mutant cRNAs injected oocytes. The above experiments were performed in mouse oocytes. Conclusion Here, we report five independent patients in which mutations in CCNB3 may be the cause of oocyte maturation arrest. Our findings shed lights on the critical role of CCNB3 in human oocyte maturation.

Untreated or inadequately treated silver−containing wastewater may pose adverse effects on hu−man health and the ecological environment. Currently, significant progress has been made in the treatment of Ag(I) in wastewater using adsorption methods, with adsorbents playing a pivotal role in this process. This paper provides a systematic review of various adsorbents for the recovery and treatment of Ag(I) in wastewater, including MOFs, COFs, transition metal sulfides, metal oxides, biomass materials, and other polymeric materials. The adsorption mechanisms of these materials for Ag(I) are elaborated upon, along with the challenges currently faced. Furthermore, insights into optimizing adsorbents and developing novel adsorbents are proposed in this study.

To solve the problem of large fluctuation and instability of photovoltaic power generation, a deep learning prediction model (TPA-TCN) based on temporal pattern attention mechanism (TPA) and temporal convolutional network (TCN) is proposed, and then applied to photovoltaic power generation. First of all, the k-means clustering algorithm is used to cluster historical data to obtain three typical weather types, and the model is trained by dividing test sets according to the clustering results. After TPA is introduced into the TCN model, which can capture the influence of each variable on the predicted series of the model, help the model pay better attention to the key features in the time series, improve the model’s ability to understand the data, and thus efficiently and accurately predict the short-term photovoltaic power. Combined with the measured data, the experiment results show that the TPA-TCN model has good generalization ability and high precision in different weather types.

The orthogonal experiment method was used to investigate the heat transfer performance of the gravity heat pipe under the influence of multiple factors such as types of working fluid, filling rate, and the aspect ratio. The Multi-index weighting method and rescaled range analysis were introduced to analyze the experimental results to propose an appropriate combination scheme of factors for the best heat transfer performance of the gravity heat pipe. The results show that gravity heat pipe can effectively reduce the internal temperature of the coal pile, thereby inhibiting the increase of the temperature of the coal pile. Moreover, the cooling capacity and effective pore radius of gravity heat pipe are positively correlated with the temperature of the coal pile. Specifically, when the coal pile temperature is about 70 ℃, the most appropriate combination of various factors is as follows, the working fluid is Al2O3-water nanofluid, the liquid filling rate is 15%, and the aspect ratio is 25.0. When the temperature of the coal pile rose to 210 ℃, the most appropriate combination of each factor is as follows, the working fluid is Al2O3-water nanofluid, the liquid filling rate is 25%, and the aspect ratio is 25.0.

Water flooding technology is an important measure to enhance oil recovery in oilfields. Understanding the pore-scale flow mechanism in the water flooding process is of great significance for the optimization of water flooding development schemes. Viscous action and capillarity are crucial factors in the determination of the oil recovery rate of water flooding. In this paper, a direct numerical simulation (DNS) method based on a Navier–Stokes equation and a volume of fluid (VOF) method is employed to investigate the dynamic behavior of the oil–water flow in the pore structure of a low-permeability sandstone reservoir in depth, and the influencing mechanism of viscous action and capillarity on the oil–water flow is explored. The results show that the inhomogeneity variation of viscous action resulted from the viscosity difference of oil and water, and the complex pore-scale oil–water two-phase flow dynamic behaviors exhibited by capillarity play a decisive role in determining the spatial sweep region and the final oil recovery rate. The larger the viscosity ratio is, the stronger the dynamic inhomogeneity will be as the displacement process proceeds, and the greater the difference in distribution of the volumetric flow rate in different channels, which will lead to the formation of a growing viscous fingering phenomenon, thus lowering the oil recovery rate. Under the same viscosity ratio, the absolute viscosity of the oil and water will also have an essential impact on the oil recovery rate by adjusting the relative importance between viscous action and capillarity. Capillarity is the direct cause of the rapid change of the flow velocity, the flow path diversion, and the formation of residual oil in the pore space. Furthermore, influenced by the wettability of the channel and the pore structure’s characteristics, the pore-scale behaviors of capillary force—including the capillary barrier induced by the abrupt change of pore channel positions, the inhibiting effect of capillary imbibition on the flow of parallel channels, and the blockage effect induced by the newly formed oil–water interface—play a vital role in determining the pore-scale oil–water flow dynamics, and influence the final oil recovery rate of the water flooding.