A5079563395- Earthquake Detection and Analysis
- Seismology and Earthquake Studies
- earthquake and tectonic studies
- Seismic Waves and Analysis
- Embedded Systems Design Techniques
- VLSI and Analog Circuit Testing
- CCD and CMOS Imaging Sensors
- RFID technology advancements
- Energy Harvesting in Wireless Networks
- Diabetic Foot Ulcer Assessment and Management
- Stroke Rehabilitation and Recovery
- Advanced Memory and Neural Computing
- Low-power high-performance VLSI design
- Wireless Body Area Networks
- Interconnection Networks and Systems
- Advancements in Semiconductor Devices and Circuit Design
- Advancements in PLL and VCO Technologies
- Knowledge Management and Technology
- Time Series Analysis and Forecasting
- Prosthetics and Rehabilitation Robotics
- Spectroscopy Techniques in Biomedical and Chemical Research
- Advanced Optical Sensing Technologies
- Optimization and Search Problems
- Spectroscopy and Chemometric Analyses
- Geophysical and Geoelectrical Methods
Peking University
2012-2023
Shenzhen MSU-BIT University
2021-2023
Peking University Shenzhen Hospital
2016-2023
Shenzhen Technology University
2023
Southwest Petroleum University
2020
IBM (United States)
2007
Nanyang Technological University
2003
ABSTRACT Earthquake forecasting is one of the most challenging tasks in field seismology that aims to save human life and mitigate catastrophic damages. We have designed a real-time earthquake framework forecast earthquakes tested it seismogenic regions southwestern China. The input data are features provided by multicomponent seismic monitoring system acoustic electromagnetic AI (AETA), which recorded using two types sensors per station: (EM) geo-acoustic (GA) sensors. target location...
The influence of earthquake disasters on human social life is positively related to the magnitude and intensity earthquake, effectively avoiding casualties property losses can be attributed accurate prediction earthquakes. In this study, an electromagnetic sensor investigated assess earthquakes in advance by collecting signals. At present, mainstream comprises two methods. On one hand, most geophysicists or data analysis experts extract a series basic features from precursor signals for...
Seismic monitoring data and some geophysical often show time series with daily-periodic amplitude jump. In this kind of series, the sharp change will naturally become focus attention, but at same time, it is easy to ignore information beyond dramatic amplitude. Therefore, paper proposes a feature extraction method for data, which can help obtain important from other than changes in magnitude. By comparing importance between obtained by proposed original feature, shows that new 1.38 times...
This paper presents a new noninvasive blood glucose monitoring method based on four near infrared spectrums and double artificial neural network analysis. We choose wavelengths, 820 nm, 875 945 1050 as transmission spectrums, capture fingers PPG signals simultaneously. The wavelet transform algorithm is used to remove baseline drift, smooth extract eight eigenvalues of each signal. are the input parameters analysis model. Double regression combines classification recognition with prediction...
With the continuous development of human society, earthquakes are becoming more and dangerous to production life society. Researchers continue try predict earthquakes, but results still not significant. data science, sensing communication technologies, there increasing efforts use machine learning methods earthquakes. Our work raises a method that applies big analysis algorithms prediction. All accumulated by Acoustic Electromagnetic Testing in One System (AETA). We propose multi-station...
A 32b 4-way SIMD dual-issue synergistic processor element of a CELL is developed with 20.9 million transistors in 14.8mm/sub 2/ using 90nm SOI technology. CMOS static gates implement the majority logic. Dynamic circuits are used critical areas, occupying 19% non-SRAM area. ISA, microarchitecture, and physical implementation tightly coupled to achieve compact power efficient design. Correct operation has been observed up 5.6GHz at 1.4V supply 56/spl deg/C.
Plantar pressure image is an important tool for gait analysis. In the rehabilitation training of stroke patients, it can provide effective quantitative data reference doctors to accurately assess recovery status patients and formulate targeted treatment programs. this paper, a recognition algorithm sensing proposed. Through DBSCAN (Density-Based Spatial Clustering Applications with Noise) clustering minimum circumscribed rectangular algorithm, footprints be located in feature extraction....
The Key Laboratory of Integrated Microsystems (IMS) Peking University Shenzhen Graduate School has deployed a self-developed acoustic and electromagnetics to artificial intelligence (AETA) system on large scale at high density in China comprehensively monitor collect the precursor anomaly signals that occur before earthquakes for seismic prediction. This paper constructs several classic time series non-time prediction models comparison analysis order find most suitable earthquake-prediction...
Wearable sensor devices organized in a Wireless Body Area Network (WBAN) have been widely used healthcare monitoring. However, energy efficiency and reliability are still major problems this area. An efficient media access control (MAC) protocol is promising solution. Traditional protocols too generic usually busy frequency band which not suitable for particular medical scenario. On the basis of Human Communication (HBC), MAC WBAN proposed work can avoid interference make non-RF technique...
To verify the relationship between AETA (Acoustic and Electromagnetics to Artificial Intelligence (AI)) electromagnetic anomalies local earthquakes, we have performed statistical studies on data observed at station. ensure accuracy of results, 20 stations with few missing abundant earthquake events were selected as research objects. A modified PCA method was used obtain sequence representing signal anomaly. Statistical results superposed epoch analysis indicated that 80% significant...
This paper describes the architecture and implementation of original gaming-oriented synergistic processor element (SPE) in both 90-nm 65-nm silicon-on-insulator (SOI) technology introduces a new SPE targeted for high-performance computing community. The Cell Broadband Engine™ contains eight SPEs. dual-issue, four-way single-instruction multiple-data is designed to achieve high performance per area power optimized process streaming data, simulate physical phenomena, render objects digitally....
The encoder and decoder are the keys of security error-rate for UHF RFID reader. In this paper, we present a multi-bit FM0/Miller design which has less more security. We take resource reused, no-gap-link between different data areas encoder, low power into consideration in our design. can handle ± 25% rate variation is 0. Under TSMC 0.18μm process, area 27685μm <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> 253μw. 53827μm 394μw.
The acquisition and analysis of plantar pressure can be applied to the postoperative gait rehabilitation training stroke patients, providing effective quantitative data reference for doctors evaluate patient's condition formulate targeted treatment plans. In this paper, an insole type sensing system was developed needs long-term continuous monitoring patients in sports training. has a large amount sensor units (maximum 288) high spatial resolution (1.64/cm2). Through optimized circuit...
In this work, a feature engineering framework is proposed for short-term earthquake prediction based on the Multi-Component Seismic Monitoring System (AETA)’ precursor data and historic seismic events. For typical electromagnetic disturbance (ED) diurnal periodic waveform which synchronize with sun rise set, an auto-recognition algorithm three parts designed corresponding features are extracted. To extract non-liner feature, Higuchi Fractal Dimension used ED data. The popular detection...
For the multi-component seismic monitoring system AETA's electromagnetic disturbance (ED) data, an outlier extraction method is proposed based on sample entropy. Sample entropy can reflect possibility of new information in time series. algorithm utilized for processing data to figure out sequence days. Then sliding quartile used obtain by calculating abnormal index sequence. The experiment result Yibin Ms 6.0 earthquake and superposed epoch analysis demonstrate that detect have ability earthquake.
Acoustic and electromagnetics to artificial intelligence (AETA) is a system used predict seismic events through monitoring of electromagnetic geoacoustic signals. It widely deployed in the Sichuan–Yunnan region (22° N–34° N, 98° E–107° E) China. Generally, signals AETA stations near epicenter have abnormal disturbances before an earthquake. When significant decrease or increase signal observed, it difficult quantify this change using only visual observation confirm that related upcoming...
This paper describes a data management server (the AETA server) that provides solution to the problem of large-area, high-density deployment and uninterrupted operation faced by ATEA system (Acoustic Electromagnetic Testing All in one system). The helps receive, process store massive monitoring efficiently reliably, as well manage maintain unattended stations remotely. In this sense, is designed from communication layer, aggregation storage deployed with load balancing cluster technology....
We found a kind of anomalous waveform that appeared frequently before and after the earthquake based on AETA original geoacoustic data. To capture these abnormal signals, we designed pattern recognition algorithm transformed data into eigenvalues. In this paper, selected for five stations in 2018 as research object. Through above-mentioned algorithm, got eigenvalues then contrasted analysis with seismic events around station met conditions magnitude shock time. Experiment results show, there...
In this paper, an integrated development environment (IDE), which is used to map application into a target reconfigurable operators (ReOps) array, presented. Having as input APU RTL description of application, the IDE produces configuration bitstream. The proposed supports variety ReOps array through revising architecture file including definition ReOps, interconnection segments and connection switches, well scale organization ReOps. A set benchmarks given verify flow IDE.
Once a majority of earthquakes occur without prediction, it is very likely to have huge impact on human society. To solve the worldwide challenging problem earthquake our laboratory has developed set sensory systems monitor abnormal activity geological signals before an happens in China. At present, more than 300 stations been deployed, and observation time exceeded 4 years. Based various activities collected, local correlation tracking method used capture signal anomalies earthquake, then...
There have been many studies in relationship between ultra-low frequency electromagnetic anomaly and earthquakes, while most of them judge using single feature. We propose a multi-feature detection method for AETA ULF disturbance signals based on Isolation Forest, with some feature extraction selection added. A statistical test superposed epoch analysis (SEA) is used its evaluation. The result shows that 6 12 selected stations show significant correlation signal earthquakes. further...
This paper reports an attempt to use low-frequency electromagnetic data (ED) from the Multi-component Seismic Monitoring System (AETA) in study of earthquake precursors China. The 5 AETA stations deployed within 200 km epicenter have been analyzed search for possible a strong that occurred JiuZhaiGou, (China) on 8 August 2017, with magnitude Mw = 7.0 and focal depth 20 km. By calculating correlation between signals every 2 stations, is made extract relatively weak anomaly information...
Abstract The research of the earthquake precursor signal anomaly is one main directions short-term and imminent prediction. An prediction method based on time precursory window proposed in this paper, which low-frequency electromagnetic signals collected by AETA. Firstly, model historical constructed machine learning method. used to detect whether current period precursors. Furthermore, two algorithms single-site group-site position paper. algorithm filters three or more stations within...