关于神经网络的区域地壳稳定性评价.pdfVIP

TU 458.4 A 2005 (1966 ) Evaluation of regional crust based on ANN ZHU Qing-jie, MA Ya-jie, CHEN Yan-hua (Civil Engineering and Architecture College, Hebei Polytechnic University, Tangshan 063009, China) Abstract: Because the stability of regional crust was controlled by many factors, and the nonlinear correlation among influencing factors was complicated, it was difficult to construct a mathematical model by traditional methods to evaluate the stability precisely during geological surveying, and many geological data could not be applied adequately. Based on artificial neural network (ANN), the evaluating model of regional crust stability was established, and the calculating method was investigated. The structure of evaluating model was introduced by taking example for Yishu faulted zone, and the correlative coefficient matrices were explained. In this model, topography, lithology of aquifer, underground water, geological structure and seismic activity were considered. So the regional crust stability in Yishu faulted zone was calculated as a basis of this model. Furthermore, model preferences were discussed and their influence on the calculated result was analyzed. According to the calculated results, some advice for the construction of railway in this area and further researches were proposed. Key words: regional crust; stability; ANN; model;