Array sísmico inalámbrico y de parámetros ambientales para la caracterización de precursores de actividad volcánica
- Autores: Angel David Moure García
- Directores de la Tesis: Joaquín del Río Fernández (dir. tes.), Daniel Mihai Toma (codir. tes.), María José Blanco (codir. tes.)
- Lectura: En la Universitat Politècnica de Catalunya (UPC) ( España ) en 2017
- Idioma: español
- Tribunal Calificador de la Tesis: Guillermo De Arcas Castro (presid.), Castro Gomariz (secret.), Itahiza Francisco Domínguez Cerdeña (voc.)
- Programa de doctorado: Programa Oficial de Doctorado en Ingeniería Electrónica
- Materias:
- Enlaces
- Tesis en acceso abierto en: TDX
- Resumen
The global seismic activity is extensively studied and well characterized by seismic networks focused on the seismic hazards derived from earthquakes. However, a specific study on volcano seismicity is not widespread, due to the wide variety of volcanic signals as well as the long periods of quiescence of many volcanoes. Seismic array technology emerged as a method for nuclear tests monitoring, however, it soon became an important tool for the analysis and location of volcano seismic signals thanks to a wide variety of processing techniques. From the techniques applied in array processing, in this PhD Thesis the beamforming algorithm has been chosen. In classical seismic arrays, the analog signals from each sensor are wired transmitted to the central acquisition system, resulting in a decrease of quality of the processed signal. Central node must perform signal acquisition from all sensors, establish a time control through the use of a GPS, store data and transmit them. Moreover, the volcanic monitoring equipment could be destroyed, or have depleted its power source, thus, it becomes evident the need of very low cost and high autonomy equipments, which currently do not exist.
In this way, a seismic array has been designed, consisting of a wireless seismic network that allows any array topology, removing the need of any wire connection. It is a significant step forward over existing devices, due to involved cost reduction and autonomy obtained, that allow fast deployment without the need of large investments. These equipments present high versatility, thanks to the use of the latest technologies in both hardware and software, using an open source operating system with embedded Linux where all software is free. This allows us to configure and display recorded data on real time, through a webpage and an instant messaging application, in order to be able to track the activity at any moment and from everywhere.
Array localization techniques are based on the search for the maximum coherence of the recorded signals by each sensor of the array, improving the quality of the seismic signal by coherently adding the signal of each element. This implies that the most important part in the data acquisition is to obtain the time delay of each signal accurately. Synthetic tests show that a delay in the synchronism of one millisecond implies an error of less than one degree in determining the arrival angle.
In order to increase the accuracy of time synchronization, it has been experimented in the laboratory with the IEEE 1588 PTP standard, using Xbee PRO commercial wireless modules, achieving synchronization of the order of microseconds, well above the data analysis requirements. The final result of this Thesis has allowed the development of a compact seismic acquisition system with low-power consumption (515,24 mW), low cost (at least 10% below average commercial systems), low-noise (2-bit A/D 24) and lightweight, making easier the possibility to develop a significant number of these devices.
The equipment has been checked in the laboratory through automated measurement systems and calibration procedures.
It has been also tested in different field campaigns where wireless seismic array was compared to a wired array of the IGN (as a reference), both using the MARK L4 seismic sensor (vertical component, and 1second natural period), being the signal-to-noise ratio of both equipment identical. Taking into account the tolerance of the A/D clock of the different stations in the array, it has been verified that in the worst case the synchronization error was 650 µs.
