Applications Of Data Mining In Rock Mechanics And Engineering

Rock has obvious characteristics of randomness and uncertainness. Accordingly, the application of artificial intelligence and relative techniques in rock mechanics and engineering is a significant study field. Data mining is the non-trivial process of identifying valid, novel, potentially useful, and ultimately understandable patterns in the data, which is a rising information technique with the development of artificial intelligence, database technique and statistic.Based on selective learning of data mining and analysis of characteristics of data or information in rock mechanics and engineering, some data mining algorithm models are applied to analysis problems of rock engineering and the research is combined with practical engineering projects. Relevance analysis to slope rock rheological test, Rock Mass Quality Assessment of dam foundation rock mass and displacement time series forecasting analysis to underground opening are performed by using data mining technique in this paper. Results of the study are detailed as follows:Firstly, relevance analysis is conducted to the record of the Xiangjiaba Project slope rock mass rheological test, using the technique of Information Gain of MS Analysis Services. The factors of the rheological test influencing on displacement of rock are studied to analysis relevance between displacement and those factors. The forecasting results illuminated that the chief effect to deformation is the total load and the following is effect time, so the test rock has evident rheological characteristic. In addition, the notable influence of experimental humidity to the rock deformation indicates that water affects the mechanics character of rock in great degree.Secondly, decision Tree classification model and Logistic Regression model are performed to Rock Mass Quality Assessment, based on SAS/Enterprise Miner. Compute parameter data in Rock Mass Quality Assessment of the Three Gorge Project dam foundation rock mass, and compare the results with former conclusion. The results accord with practical engineering projects. The same methods are applied to Rock Mass Quality Assessment of the Shuibuya Project dam foundation rock mass, and the rock mass fit for dam foundation is selected.Lastly, the time series analysis and time series data mining system of SAS/ETS are used to conduct tendency forecasting of displacement of the Longtan Water Power Project underground plant observation tunnel rock. According to the forecasting analysis results, the displacement of the rock mass is mainly trending to steady rheological status. Different positions of the opening has different rheologicalcharacteristic and the time of reaching steady rheological status is different: rock mass at the foot of the opening has evident accelerated rheological phenomenon.In conclusion, the above-mentioned three applications and studies are integrated with practical engineering projects, respectively referring to slope rock mass, dam foundation rock mass and underground opening rock mass. At the same time, the computing data from laboratory experiment and spot observation has project background. For these reasons, though it is not so perfect up to now, the author believe that data mining will play a more important role in rock mechanics and engineering in future.