A Real-Time Early Warning Seismic Event Detection Algorithm Using Smart Geo-Spatial Bi-Axial Inclinometer Nodes for Industry 4.0 Applications

Applied Sciences

Adel Ben Mnaouer¹  MohammedAbdulla E. Al-Hitmi²  Farid Touati²  Hasan Tariq²  Damiano Crescini³ 

[1] Department of Computer Engineering and Computational Sciences, Faculty of Engineering, Applied Sciences and Technology, Canadian University Dubai, 117781 Dubai, UAE;Department of Electrical Engineering, College of Engineering, Qatar University, 2713 Doha, Qatar;Dipartimento di Ingegneria delI’Informazione, Brescia University, 25121 Brescia, Italy;

关键词: applied methods;    earthquake;    seismic waves;    real-time detection;    early warning;    inclinometers;    Internet of Things (IoT);   

DOI  :  10.3390/app9183650

来源: DOAJ

【 摘 要 】

Earthquakes are one of the major natural calamities as well as a prime subject of interest for seismologists, state agencies, and ground motion instrumentation scientists. The real-time data analysis of multi-sensor instrumentation is a valuable knowledge repository for real-time early warning and trustworthy seismic events detection. In this work, an early warning in the first 1 micro-second and seismic wave detection in the first 1.7 milliseconds after event initialization is proposed using a seismic wave event detection algorithm (SWEDA). The SWEDA with nine low-computation-cost operations is being proposed for smart geospatial bi-axial inclinometer nodes (SGBINs) also utilized in structural health monitoring systems. SWEDA detects four types of seismic waves, i.e., primary (P) or compression, secondary (S) or shear, Love (L), and Rayleigh (R) waves using time and frequency domain parameters mapped on a 2D mapping interpretation scheme. The SWEDA proved automated heterogeneous surface adaptability, multi-clustered sensing, ubiquitous monitoring with dynamic Savitzky−Golay filtering and detection using nine optimized sequential and structured event characterization techniques. Furthermore, situation-conscious (context-aware) and automated computation of short-time average over long-time average (STA/LTA) triggering parameters by peak-detection and run-time scaling arrays with manual computation support were achieved.

【 授权许可】

Unknown