Application of pattern recognition analysis of rock acoustic emission for determination of Kaiser Effect

摘要

Nowadays, acoustic emission testing based on the Kaiser Effect is increasingly used for estimating in-situ stress in laboratories. The Kaiser effect can be observed in rocks and other material during their cyclic uniaxial loading by AE moni-toring and corresponds to a stress memory. As soon as the load achieves its previously reached peak value, a noticeable increase in acoustic emission activity takes place. Different methods have been proposed for determining the KE point by different researchers. There are cases where KE cannot be observed clearly by plotting data and using the different methods. Consequently, the KE results can be doubtful because determining the KE value mostly is difficult. In this paper, AE testing and monitoring of AE signals have been accomplished under uniaxial compressive testing on the cylindrical samples of sandstone. After each preloading stage, AE testing was carried out without delay by applying a loading stress more than the previous loading stress level. Pattern recognition is the study of how machines can observe the environment, learn to distinguish patterns of interest from their background, make sound and reasonable decisions about the categories of the patterns. It is the science of the methods of data analysis for the purpose of separating, clustering and classifying them based on a number of features. One of the most important issues in the science of pattern recognition which has recently attracted considerable attention in different sciences is combining classifiers.