|
|
|
题名
|
作者
|
年代
|
出处
|
被引量
|
| 1 | Deep learning of rock microscopic images for intelligent lithology identification: Neural network comparison and selection显示文摘An intelligent lithology identification method is proposed based on deep learning of the rock microscopic images.Based on the characteristics of rock images in the dataset,we used Xception,MobileNet_v2,Inception_ResNet_v2,Inception_v3,Densenet121,ResNet101_v2,and ResNet-101 to develop microscopic image classification models,and then the network structures of seven different convolutional neural networks(CNNs)were compared.It shows that the multi-layer representation of rock features can be represented through convolution structures,thus better feature robustness can be achieved.For the loss function,cross-entropy is used to back propagate the weight parameters layer by layer,and the accuracy of the network is improved by frequent iterative training.We expanded a self-built dataset by using transfer learning and data augmentation.Next,accuracy(acc)and frames per second(fps)were used as the evaluation indexes to assess the accuracy and speed of model identification.The results show that the Xception-based model has the optimum performance,with an accuracy of 97.66%in the training dataset and 98.65%in the testing dataset.Furthermore,the fps of the model is 50.76,and the model is feasible to deploy under different hardware conditions and meets the requirements of rapid lithology identification.This proposed method is proved to be robust and versatile in generalization performance,and it is suitable for both geologists and engineers to identify lithology quickly. | Zhenhao Xu Wen Ma Peng Lin Yilei Hua | 2022 | Journal of Rock Mechanics and Geotechnical Engineering2022，14，4： | 2 |
| 2 | Adverse Geology Identification Through Mineral Anomaly Analysis During Tunneling:Methodology and Case Study显示文摘Accurate and effective identification of adverse geology is crucial for safe and efficient tunnel construction.Current methods of identifying adverse geology depend on the experience of geologists and are prone to misjudgment and omissions.Here,we propose a method for adverse geology identification in tunnels based on mineral anomaly analysis.The method is based on the theory of geoanomaly,and the mineral anomalies are geological markers of the presence of adverse geology.The method uses exploration data analysis(EDA)to calculate mineral anomaly thresholds,then evaluates the mineral anomalies based on the thresholds and identifies adverse geology based on the characteristics of the mineral anomalies.We have established a dynamic expansion process for background samples to achieve the dynamic evaluation of mineral anomalies by adjusting anomaly thresholds.This method has been validated and applied in a tunnel excavated in granite.As shown herein,in the tunnel range of 142+800–142+860,the fault F37 was successfully identified based on an anomalous decrease in the diagenetic minerals plagioclase and hornblende,as well as an anomalous increase in the content of the alteration minerals chlorite,laumonite,and epidote.The proposed method provides a timely warning when a tunnel enters areas affected by adverse geology and identifies whether the tunnel is gradually approaching or moving away from the fault.In addition,the applicability,accuracy,and further improvement of the method are discussed.This method improves our ability to identify adverse geology,from qualitative to quantitative,and can provide reference and guidance for the identification of adverse geology in mining and underground engineering. | Zhenhao Xu Tengfei Yu Peng Lin Shucai Li | 2023 | Engineering2023，，8： | 0 |
| 3 | Extraction and imaging of indicator elements for non-destructive,in-situ,fast identification of adverse geology in tunnels显示文摘The lag in quantitative methods and detection techniques for geologic information has resulted in time-consuming and human-experienced geologic analysis in tunnels.Geochemical indicators of rocks can be used to identify adverse geology and to explain the intrinsic causes of damage to normal rocks.This study proposes a method to identify adverse geology by extracting and imaging the indicator elements.The mapping relationship between rock components and geologic bodies is quickly determined by indicator element extraction based on factor analysis,and then the data are gridded for image output.The location and size of the target adverse geology are visually identified through the distribution images of the indicator elements,thus reducing data dimensions and analysis time.A non-destructive,in-situ and fast element detection technique in tunnels was adopted to speed up the process of geology identification.The accuracy of the detection was validated by comparing field and laboratory test results.This study further confirms and refines the previous research,and the results provide references for geological,mining and underground projects. | Fumin Liu Peng Lin Zhenhao Xu Ruiqi Shao Tao Han | 2023 | International Journal of Mining Science and Technology2023，33，12： | 0 |
| 4 | Spin injection into heavily-doped n-GaN via Schottky barrier显示文摘Spin injection and detection in bulk GaN were investigated by performing magnetotransport measurements at low temperatures.A non-local four-terminal lateral spin valve device was fabricated with Co/GaN Schottky contacts.The spin injection efficiency of 21%was achieved at 1.7 K.It was confirmed that the thin Schottky barrier formed between the heavily ndoped GaN and Co was conducive to the direct spin tunneling,by reducing the spin scattering relaxation through the interface states. | Zhenhao Sun Ning Tang Shuaiyu Chen Fan Zhang Haoran Fan Shixiong Zhang Rongxin Wang Xi Lin Jianping Liu Weikun Ge Bo Shen | 2023 | Journal of Semiconductors2023，44，8： | 0 |
| 5 | Investigation on the Effect of the Multilayered Porous Structure of Sea Urchin Skeleton on Its Mechanical Behavior显示文摘In this paper,the effect of stereom structure on the mechanical behavior of the Sea Urchin Inorganic Skeleton(SUIS)has been studied.The stereom microstructure of both Anthocidaris crassispina and Tripnenstes gratilla was characterized by Scanning Electron Microscopy(SEM).Results indicate that a three-layer porous structure consisting of a growth,a support,and a resorption(GSR)layer is a common denominator for both species.The effect of GSR layer order on the mechanical behavior of the SUIS was studied by a finite element method.The results show that the GSR model could effectively reduce the maximum tensile stress on its meridional sutures under unidirectional pressure,hydrostatic pressure,and self-weight situation.For a fabricated three-layered ceramic test strips with different layer orders,the mechanical properties have a completely opposite performance compared with the compressive properties of the calculated SUIS-Iike models.This indicates that the GSR structure can effectively improve the mechanical properties of the SUIS,but it cannot be applied to bionics without considering its synergistic effect with the macro-structure of the SUIS.This is a typical example of bionic invalidation by single structure,where multi-level structure bionics may be an effective solution. | Hui Yu Jianbao Li Zhenhao Hou Jianlin Li Yongjun Chen Chunfu Lin | 2020 | Journal of Bionic Engineering2020，17，3： | 0 |
| 6 | Crop growth inhibited by over-liming in tea plantations显示文摘Liming is a common strategy applied to attain optimal pH for tea growth in severely acidic soils.Tea however is a calciphobous plant,and the effects of liming on its growth and nutrient uptake remain poorly understand.Therefore,it is necessary to better understand the effects of liming on soil chemical properties and tea nutrient content.In this study,a tea plantation that had exhibited large variation in growth after liming was selected as a study site.We categorized plots into two growth condition groups:Plot 1(poor growth)and Plot 2(excellent growth).Tea nutrient levels,and soil chemical properties were then compared between the two groups.Normalized difference vegetation index(NDVI)and transformed vegetation index(TVI)values were significantly higher and lower,respectively,in Plot 2 than in Plot 1.Yield,number of buds per m2,and 100-bud weight were significantly higher in Plot 2.These results were attributed to higher N,K,and Al concentrations and lower Ca concentrations in leaves,and lower pH and available Ca and higher available Al in soil.Leaf concentrations of K and Al were significantly negatively correlated with leaf concentrations of Ca and soil pH.A positive relationship was observed between leaf concentrations of K and Al,indicating inhibited K and Al uptake due to over-liming,restricting tea growth.In conclusion,our results show that tea growth will be restricted by over-liming,as a result of the high soil pH and Ca concentration inhibiting the K and Al uptake. | Peng Yan Zhenhao Zou Jingcheng Zhang Lin Yuan Chen Shen Kang Ni Yeliang Sun Xin Li Liping Zhang Lan Zhang Jianyu Fu Wenyan Han | 2021 | Beverage Plant Research2021，1，1： | 0 |
| 7 | A Method of Text Extremum Region Extraction Based on Joint-Channels显示文摘Natural scene recognition has important significance and value in the fields of image retrieval,autonomous navigation,human-computer interaction and industrial automation.Firstly,the natural scene image non-text content takes up relatively high proportion;secondly,the natural scene images have a cluttered background and complex lighting conditions,angle,font and color.Therefore,how to extract text extreme regions efficiently from complex and varied natural scene images plays an important role in natural scene image text recognition.In this paper,a Text extremum region Extraction algorithm based on Joint-Channels(TEJC)is proposed.On the one hand,it can solve the problem that the maximum stable extremum region(MSER)algorithm is only suitable for gray images and difficult to process color images.On the other hand,it solves the problem that the MSER algorithm has high complexity and low accuracy when extracting the most stable extreme region.In this paper,the proposed algorithm is tested and evaluated on the ICDAR data set.The experimental results show that the method has superiority. | Xueming Qiao Weiyi Zhu Dongjie Zhu Liang Kong Yingxue Xia Chunxu Lin Zhenhao Guo Yiheng Sun | 2020 | Journal on Artificial Intelligence2020，2，1： | 0 |
| 8 | Solid-liquid flow characteristics and sticking-force analysis of valve-core fitting clearance显示文摘External contamination particles or wear particles corroded by a valve body are mixed into the fluid. As a result, when the fluid enters the fitting clearance of the valve core, it can cause an increase in resistance and lead to sticking failure of the valve core. This paper analyzes solid-liquid flow characteristics in fitting clearances and valve-core sticking based on the Euler-Euler model, using a typical hydraulic valve as an example. The impact of particle concentration and diameter on flow characteristics and valve-core sticking force was analyzed. The highest volume fraction of particles was in the pressureequalizing groove(PEG), with peak values increasing as the particle diameter increased. The sticking force increased with increasing particle concentration. When the particle diameter was 12 μm, the sticking force was the largest, making this the sensitive particle diameter. Particle distribution and valve-core sticking force were compared for oval, rectangular, and triangular PEGs. The fluid-deflection angles in oval and rectangular PEGs were larger, and their values were 32.83° and 39.15°, respectively. The fluid-deflection angle in the triangular PEG was relatively small, less than 50% that of the oval or rectangular PEGs. The particle-volume-fraction peaks in oval, rectangular, and triangular PEGs were 0.0317, 0.0316, and 0.0312, respectively. The sticking forces of oval, rectangular, and triangular PEGs were 4.796, 4.802, and 4.757 N, respectively when the particle diameter was 12 μm. This work provides a reference for design and research aimed at reducing valve-core sticking. | Jin-yuan QIAN Jiaxiang XU Fengping ZHONG Zhenhao LIN Tingfeng HUA Zhijiang JIN | 2023 | Journal of Zhejiang University-Science A(Applied Physics & Engineering)2023，24，12： | 0 |
您所在的机构暂未开通此部分的功能，请联系管理员