En  |  Pyc

Perimeter seismic monitoring


In 2004 under preliminary agreement with Russian Oil and Gas transporting company SYNAPSE started SNDA redesign in order to create real-time perimeter protection product working on the computers with restricted computational and power resources. In 2005 SYNAPSE performed first field demonstration of the system, capable to provide remote (IEEE-802.11b) monitoring of the 300 meter length frontier equipped with a geophone string comprising only 3 (three) 4.5-Hertz sensors. Raw waveform data from the digitizer (data concentrator) buried in a ground was transmitted to the wireless laptop of the pipeline patrol, where pattern recognition software (SNDA new library set) discriminated between vehicle, man, digging and environmental noise. The experiment was conducted just 200 meters away from the extremely loaded highway (Moscow Ring Road), in a forest with relatively windy weather (wind speed 4-5 m/s), along the lake shore, 10 to 200 meters to the brook.

The recognition results were excellent and new agreement was signed for the second system prototype development capable to provide mobile telephone notification of the captured intruder. This work was finished by spring 2006. At the same time new cooperative project between SYNAPSE and Telecommunication Academy started on porting SYNAPSE software to digital signal processor environment and developing a portable low power high resolution 24 bit digital seismic sensor with wireless notification capability. The sensor will have powerful data processing algorithms, relatively low power consumption and wide temperature range. Obviously it requires a lot of engineering and scientific efforts to produce the equipment responding to such contradictory requirements. SYNAPSE continued research and algorithms modifications to facilitate object recognition in noisy environments. Also, extensive research was provided in the area of data collection and algorithms learning. We continued this work with the database of the following signals: human steps, underground drilling and other anthropomorphic activity, vehicles, trucks, planes, helicopters, animals, and different background noise.

This work is partially based on the research we conducted within different SNDA projects for low energy local seismic events locating and discriminating. SNDA includes a lot of data processing libraries, such as different filtering, F-K analysis, emission tomography, beamforming, wavelet analysis, artificial neural networks, and others, allowing obtaining good results working with "very close objects" from the point of view of observational seismology. We are encouraged with the results obtained at the western Siberia oilfields while locating the well bore at the depths up to 2500 meters with the arrays of seismic sensors deployed hundreds of meters away from the well. It will allow to apply this technique for different other civil and defense monitoring techniques, for instance underground drilling detection.

The developed systems will be applied in different areas. First of all, this is anti-criminal and anti-terrorist pipeline monitoring. Another application - pipeline protection during the construction works, when heavy mechanisms can damage pipes buried in a ground. Separate software version is under development for specific perimeter guarding application with the set of geophysical sensors, including magnetic and acoustic.

Your are here: