TY - JOUR
T1 - Passive seismic monitoring in conventional tunnelling
T2 - An innovative approach for automatic process recognition using support vector machines
AU - Hartl, Irene
AU - Sorger, Marcel
AU - Hartl, Karin
AU - Ralph, Benjamin
AU - Schlögel, Ingrid
PY - 2023/4/14
Y1 - 2023/4/14
N2 - For the underground construction sector as in conventional tunnelling, there is still a lack of automatization and digitalization progresses, especially concerning tunnel construction monitoring. A manual documentation of the time intervals for subsequent processes by the respective employee is currently state of the art. This study addresses a cost and time effective data acquisition and evaluation method using conventional geophones for the differentiation of the processes involved in tunnel construction by analysis of elastic wave signals. The field experiments were executed at the construction site of “Zentrum am Berg” in Austria where seismic signals were recorded during the conventional tunnel excavation process. The seismic emissions induced by the respective machinery during different constructuon steps are distinguished with a machine learning approach using support vector machines, leading to the possibility of associating them with the corresponding time of the machinery in use. The semi-automatic evaluation of the gathered data should facilitate the documentation of the daily working diagrams, supplement project management and effective planning and optimize predictive maintenance possibilities in the underground construction industry.
AB - For the underground construction sector as in conventional tunnelling, there is still a lack of automatization and digitalization progresses, especially concerning tunnel construction monitoring. A manual documentation of the time intervals for subsequent processes by the respective employee is currently state of the art. This study addresses a cost and time effective data acquisition and evaluation method using conventional geophones for the differentiation of the processes involved in tunnel construction by analysis of elastic wave signals. The field experiments were executed at the construction site of “Zentrum am Berg” in Austria where seismic signals were recorded during the conventional tunnel excavation process. The seismic emissions induced by the respective machinery during different constructuon steps are distinguished with a machine learning approach using support vector machines, leading to the possibility of associating them with the corresponding time of the machinery in use. The semi-automatic evaluation of the gathered data should facilitate the documentation of the daily working diagrams, supplement project management and effective planning and optimize predictive maintenance possibilities in the underground construction industry.
U2 - 10.1016/j.tust.2023.105149
DO - 10.1016/j.tust.2023.105149
M3 - Article
SN - 0886-7798
VL - 137.2023
JO - Tunnelling and Underground Space Technology
JF - Tunnelling and Underground Space Technology
IS - July
M1 - 105149
ER -