Research On Acoustic Emission Signal Of Rock Burst Based On Data Mining

Natural rock contains a large number of initial defects,the water content of rocks or not and how much their water content have a significant impact on the stability of rock engineering.With the continuous development of acoustic emission monitoring technology and uniaxial loading acoustic emission technology,the research on the stability of water-bearing rocks is also deepening.Rocks with different water contents will also have different numbers of cracking events under different acoustic loadings.With the release of different energy.Analyzing the acoustic emission data and acoustic emission(AE)energy data of water-bearing rocks is helpful to predict the rock moisture content and rock instability and improve the safety of rock engineering.Based on the uniaxial loading acoustic emission test,three kinds of clustering methods,k-means clustering method,fuzzy C-means clustering method and subtractive clustering method,were used to conduct the one-and two-dimensional Class analysis,made some research results.Three kinds of clustering methods were used to cluster the number of AE events and AE energy values one by one.The number of events and the amount of energy were defined.The one-dimensional clustering results of two parameters are combined to get the change of event number and energy in the entire process of rock failure.At the same time,it was found that the number of events before the rock rupture increased dramatically and could be used as the rupture precursor information.Three kinds of clustering methods were used to cluster the acoustic emission events and AE energy values in two dimensions.The relationship between the number of events and the corresponding energy in the process of rock rupture was concluded.Some conclusions were drawn that the release of large energy was some small energy The accumulation of a few large energy release,they appear in different stages of the rock rupture.