A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
An automatic classification method for microseismic events and blasts during rock excavation of underground caverns
Highlights Proposed a method to automatically extract the features of microseismic and blasting signals. Adopted an effective machine learning algorithm for automatical classification. Efficient recognition of microseismic signals with little operator interference. Offered more accurate microseismic data for source inversion of rock failure.
Abstract Accurately acquiring microseismic (MS) signals is the cornerstone of MS monitoring during underground rock excavation. This study developed a classification method to automatically recognize MS signals under the interference of blasting signals during construction period. Our proposed method consists of the improved complete ensemble empirical mode decomposition with adaptive noise (I-CEEMDAN), singular value decomposition (SVD) and k-nearest neighbors algorithm (k-NN). I-CEEMDAN was taken to decompose original multi-frequency to a few mono-frequency signal subcomponents and SVD was adopted to extract singular values from matrices formed by the decomposed results. These obtained singular values, regarded as input features, were imported into k-NN to establish an automatic classification model to identify MS signals. The 500 signals collected from Wudongde hydropower station in Southwest China were analyzed using our method. Our numerical experiments indicated that I-CEEMDAN and SVD can extract key characteristics of the signals, and k-NN has higher identification accuracy and computational efficiency compared with other machine learning algorithms. Our proposed method can be applied in MS monitoring techniques to offer more accurate MS signals for subsequent source analysis to achieve disasters warning.
An automatic classification method for microseismic events and blasts during rock excavation of underground caverns
Highlights Proposed a method to automatically extract the features of microseismic and blasting signals. Adopted an effective machine learning algorithm for automatical classification. Efficient recognition of microseismic signals with little operator interference. Offered more accurate microseismic data for source inversion of rock failure.
Abstract Accurately acquiring microseismic (MS) signals is the cornerstone of MS monitoring during underground rock excavation. This study developed a classification method to automatically recognize MS signals under the interference of blasting signals during construction period. Our proposed method consists of the improved complete ensemble empirical mode decomposition with adaptive noise (I-CEEMDAN), singular value decomposition (SVD) and k-nearest neighbors algorithm (k-NN). I-CEEMDAN was taken to decompose original multi-frequency to a few mono-frequency signal subcomponents and SVD was adopted to extract singular values from matrices formed by the decomposed results. These obtained singular values, regarded as input features, were imported into k-NN to establish an automatic classification model to identify MS signals. The 500 signals collected from Wudongde hydropower station in Southwest China were analyzed using our method. Our numerical experiments indicated that I-CEEMDAN and SVD can extract key characteristics of the signals, and k-NN has higher identification accuracy and computational efficiency compared with other machine learning algorithms. Our proposed method can be applied in MS monitoring techniques to offer more accurate MS signals for subsequent source analysis to achieve disasters warning.
An automatic classification method for microseismic events and blasts during rock excavation of underground caverns
Jiang, Ruochen (author) / Dai, Feng (author) / Liu, Yi (author) / Wei, Mingdong (author)
2020-04-18
Article (Journal)
Electronic Resource
English
Microseismic Monitoring Around Large Underground Storage Caverns During Construction
| British Library Conference Proceedings | 2006
Microseismic event monitoring of highly stressed rock mass around underground oil storage caverns
| Online Contents | 2006
Microseismic event monitoring of highly stressed rock mass around underground oil storage caverns
| British Library Online Contents | 2006