Geo-stress field fitting based on the partial least square method
-
摘要: 采用偏最小二乘回歸分析方法分析礦區地應力場分布.以烏鞘嶺隧道嶺脊地段的有限測點的地應力測量結果為樣本,使用商業有限元軟件ANSYS對嶺脊地段區域的地應力場進行數值模擬.利用地應力實測數據與計算結果進行偏最小二乘回歸計算,得到擬合的地應力場數據.與傳統最小二乘分析方法相比,本文的方法通過主成分分析,可以有效解決多因變量回歸計算時由于響應變量之間較高的相關性而導致的回歸計算誤差,并可以有效地反映變量間的關系,為地應力場的建立提供有效的數據支持.Abstract: Geo-stress distribution in a mining area was analyzed by means of a partial least squares regression method (PLSRM). Based on data measured in Wushaoling tunnel ridge section, a numerical simulation of the initial geo-stress field around the area was performed using the commercial software ANSYS. Partial least squares regression analysis was carried out through actual measurement data and calculation results of geo-stress, and then fitting data was ready. Compared with the traditional least squares method, due to the advantage of principal component analysis in applying PLSRM, the present method decreases the regression error caused by the correlation between response variables. It can more effectively reflect the features of discontinuous geo-stress distribution and thus provides effective data support for establishing the geo-stress field.
-
Key words:
- partial least squares /
- geo-stress field /
- correlation /
- regression analysis
-

計量
- 文章訪問數: 267
- HTML全文瀏覽量: 108
- PDF下載量: 5
- 被引次數: 0