Yunxiang Liu

Shanghai Institute of Technology

Author Statistics

Papers

Citation

H-Index

i-10 index

Research Trends

Author Order

Document Type

Co-Authors

Y. Jade Morton

University of Colorado Boulder

Zhe Yang

Tongji University

Zhanxiang He

Southern University of Science and Technology

Zuzhi Hu

China National Petroleum Corporation (China)

Wenju Zhao

China National Petroleum Corporation (China)

Hualu Si

China National Petroleum Corporation (China)

Lei Liu

University of Colorado Boulder

Shuang Liu

China University of Geosciences

Xiangyun Hu

China University of Geosciences

Xiange Jian

Bureau of Geology and Mineral Exploration and Development of Guizhou Province

Cooperative Institutions

China National Petroleum Corporation (China)

Chinese Academy of Sciences

China University of Geosciences (Beijing)

University of Chinese Academy of Sciences

Henan University

Schlumberger (British Virgin Islands)

Ocean University of China

Zhejiang University

Sichuan University

Northwest A&F University

Author Statistics

Papers

Citation

H-Index

i-10 index

Research Field

Machine Learning Prediction of Highlatitude Ionospheric Irregularities from GNSS-derived ROTI Maps

Proceedings of the Satellite Division's International Technical Meeting (Online)/Proceedings of the Satellite Division's International Technical Meeting (CD-ROM) (2021)

Lei Liu Y. Jade Morton Yunxiang Liu

This paper applies a convolutional long short-term memory (convLSTM) model with a custom- designed loss function Lc (convLSTM-Lc) to forecast high-latitude ionospheric irregularities. The data used for this study are GNSS-derived rate of TEC change index (ROTI) maps collected in 2014-2016. The convLSTM-Lc is trained with real ROTI data above 51°magnetic latitude (MLat) during January 1, 2014 to October 21, 2015. Test results show that the convLSTM-Lc model can predict large-scale features and overall evolution of the high-latitude ionospheric irregularities occurrence with a lead time of up to one day.

10.33012/2021.18046

Cite

Citations (1)

Datasets for 'Time Lags between Ionospheric Scintillation Detection at Northern Auroral Latitudes and Onset of Geomagnetic Storms'

Zenodo (CERN European Organization for Nuclear Research) (2023)

Zhe Yang Y. Jade Morton Yunxiang Liu

Interplanetary scintillation

10.5281/zenodo.8088608

Cite

Citations (0)

Machine Learning Prediction of Storm‐Time High‐Latitude Ionospheric Irregularities From GNSS‐Derived ROTI Maps

Geophysical Research Letters (2021)

Lei Liu Y. Jade Morton Yunxiang Liu

Abstract This study presents an image‐based convolutional long short‐term memory (convLSTM) machine learning algorithm to predict storm‐time ionospheric irregularities. Unlike existing methods that are either focused on irregularities at individual locations or treat the irregularity prediction as a classification problem, the convLSTM‐based architecture forecasts an entire regions' ionospheric irregularity occurrence and intensity values. We implemented the convLSTM‐based model with a custom‐designed loss function (convLSTM‐ L c ) that includes a dynamic penalty on the difference between the truth and the predicted rate of total electron content index (ROTI) maps. The convLSTM‐ L c is trained with real ROTI data collected during January 1–August 7, 2015 from ∼550 global navigation satellite system receivers located in (45°–90°N, 0°–180°W). Test results show that the convLSTM‐ L c algorithm can forecast irregularity structures more accurately than a convLSTM model that implements conventional loss functions. The model also outperforms the convLSTM‐ L 1 , convLSTM‐ L 2 , and persistence models according to statistical classification metrics with a lead time of up to 60 min.

10.1029/2021gl095561

Cite

Citations (26)

Indoor 3D wireframe construction from incomplete point clouds based on Gestalt rules

International Journal of Applied Earth Observation and Geoinformation (2024)

Zhiqiang Qin Xiaoli Liang Jiayao Wang Xianjun Gao L. Chen

The complexity of indoor point cloud and information loss during point cloud acquisition are the main challenges for accurate and complete indoor buildings 3D wireframe construction. This paper presents a 3D wireframe construction method for indoor structure point clouds based on Gestalt rules. First, boundary points of indoor structures are used for structural plane segmentation, including ceilings, floors, walls, beams, etc. Then, point clouds of each structural plane are projected onto its best-fit plane to obtain a 2D projection, followed by boundary points extraction and line fitting to get the 2D line segments. Finally, an accurate and complete 3D wireframe is constructed by optimizing anomalous line segments based on the Gestalt rules. In our method, Gestalt rules are used to deal with problems in line segments, including cluttered distribution, duplicate extraction, non-regularization (non-parallel and non-perpendicular), topological errors (non-closed and discontinuous), and missing information. Experiments demonstrate that our proposed method significantly improves the accuracy and completeness of the results.

10.1016/j.jag.2024.103893

Cite

Citations (1)

Topology optimization design for offshore platform jacket structure

Applied Ocean Research (2019)

Xiaojie Tian Qingyang Wang Guijie Liu Yunxiang Liu Yingchun Xie

Topology optimization

10.1016/j.apor.2019.01.003

Cite

Citations (57)

Anti-interference Test for the New SEP Instrument: CSAMT Study at Dongguashan Copper Mine, China

Journal of Environmental and Engineering Geophysics (2017)

Da Lei Qingyun Di Junjie Wu Xingchun Wang Yunxiang Liu

In this paper, an independently developed device system called Surface Electromagnetic Prospecting (SEP) system was introduced through a CSAMT test at Dongguashan copper mine, east of Tonglin, Anhui province. In this area exists a strong electromagnetic interference including mineral and other human interferences, so there is a big challenge for electromagnetic exploration field work. In order to test the anti-interference ability of our system and ensure the quality of data, we applied both spectral analysis and temporal filter methods to improve the signal-to-noise ratio. After data processing, we obtained better results in our geophysical models and showed that the SEP system is able to obtain stable and reliable data in a complex and noisy environment. Consequently, the anti-interference capability of the SEP system is capable to undertake complicated exploration tasks.

Prospecting

Data Processing

10.2113/jeeg22.4.339

Cite

Citations (15)

Machine Learning-based Investigation of Feature Importance for High-latitude Ionospheric Scintillation Forecasting

Proceedings of the Institute of Navigation ... International Technical Meeting/Proceedings of the ... International Technical Meeting of The Institute of Navigation (2021)

Alexis J. Wu Yunxiang Liu

This paper introduces a machine learning approach to investigate the feature importance for scintillation forecasting. Here, the features are historical measurements used as input for machine learning models. We propose to use gradient boosting as the machine learning algorithm to conduct a scintillation forecasting task at high latitudes. Once the gradient boosting model is trained, the rank of feature importance can be obtained. The preliminary results show that the top 10 most important features indeed are correlated with the future occurrence of scintillation. The feature importance ranking has the potential to guide feature selection for machine learning-based scintillation forecasting and improve forecasting performance. In addition, the feature importance list could also provide insights on the investigation of the complex coupling between solar wind and ionospheric disturbance.

Gradient boosting

Feature (linguistics)

Interplanetary scintillation

Boosting

10.33012/2021.17855

Cite

Citations (2)

Nonseismic methods guide seismology in steep structures

The Leading Edge (2002)

Zhanxiang He Yunxiang Liu Lunkai Yang

The Mashan structural belt, which extends several hundred km east-west and 8–12 km north-south, is in the east section of the southern faulted tectonic belt at Bachu uplift of Tarim Basin. It is mainly controlled by two large faults—one in the south and one in the north. Commercial gas has been discovered in the area and it is considered to have further exploration potential despite the huge obstacles its topography presents to seismic prospecting. Seismic data quality in the south Mashan is poor and the reflection from top Ordovician limestone, the exploration target, is not clear. Traditionally, the Mashan structural belt was regarded as a single anticline with a northern high as the principal feature (Figure 1). However, comprehensive nonseismic geophysical studies (gravity, magnetic, and electromagnetic methods) have resulted in a different view—one that has attracted oil explorationists. Figure 1. Seismic interpretation profile of the Mashan tectonic belt. The Mashan structural belt was formed during Himalayan orogeny. Drilling data show that the Quaternary and Tertiary consist of sandstone, shale, and lagoon sediments; the upper Permian mainly is sandstone-shale; the lower Permian has a basalt section and shale section; the Carboniferous is the continental-oceanic interaction sediment; and the Ordovician mainly consists of the marine sediment. Based on core density measurements and the P-wave interval velocity, we conclude that the interface between the Permian-Carboniferous and Ordovician (−0.15g/cm3) and the interface between the Cenozoic and Permian-Carboniferous (nearly −0.15g/cm3) are the principal density interfaces in this area. The lower Permian basalt has stronger magnetism but its thickness is no more than 200–350 m. Basement metamorphic rock has middle or weak …

10.1190/1.1508948

Cite

Citations (1)

Determination of the magnetization direction via correlation between reduced‐to‐the‐pole magnetic anomalies and total gradient of the magnetic potential with vertical magnetization

Geophysical Prospecting (2024)

Xiange Jian Shuang Liu Zuzhi Hu Yunxiang Liu Hongzhu Cai

Abstract The total magnetization of an underground magnetic source is the vector sum of the induced magnetization and the natural remanent magnetization. The direction of the total magnetization serves as important a priori information in the inversion and processing of magnetic data. We demonstrated that the total gradient of the magnetic potential with vertical magnetization constitutes the envelope of the vertical component of the magnetic field for all directions of the Earth's field and source magnetization. The total gradient of the magnetic potential with vertical magnetization and the reduction‐to‐the‐pole field simultaneously tend to achieve maximum symmetry near the correct total magnetization direction. As a result, the total magnetization direction can be estimated by computing the correlations between the reduction‐to‐the‐pole and the total gradient of the magnetic potential with vertical magnetization. The proposed method yields accurate magnetization directions in synthetic model examples. The total gradient of the magnetic potential with vertical magnetization is less susceptible to data noise than transforms which are derived from the high‐order magnetic field derivatives or tensors. The estimation results are slightly affected by changes in the source magnetization direction. In a field example in the Weilasito region (North China), the reduction‐to‐the‐pole fields calculated using the estimated magnetization directions are well centred over the source. The proposed method obtained a more focused magnetization direction than that of a three‐dimensional magnetization vector inversion. The total gradient of the magnetic potential with vertical magnetization therefore provides a novel and accurate approach to determine the total magnetization direction from the total field anomaly in a variety of situations.

Stoner–Wohlfarth model

Orbital magnetization

Single domain

10.1111/1365-2478.13499

Cite

Citations (0)

The study of 3D gravitational and magnetic data acquisition, processing, interpretation and its application

Yunxiang Liu XU Xiao-fang Hualu Si

Gravitational and magnetic surveys are important exploration methods for oil and gas. With increasing complicity of exploration targets, 3D gravity and magnetic exploration will play a more important role than the conventional gravity and magnetic surveys. 3D gravitational and magnetic exploration techniques include 3D acquisition, processing and inversion. Of these, the 3D inversion technique is key. 3D gravitational and magnetic techniques are expected to play an important role in studying deep‐seated objectives and detailed structure interpretation.

10.1190/1.3603798

Cite

Citations (1)