Based on existing methods eight common parameters affecting the soil liquefaction are considered and weighted through the principal component analysis. To reduce the dimensionality of the data set, four principal components are obtained by projecting the multivariate data vectors on the space spanned by the eigenvectors. Using the available field liquefaction and non liquefaction data, the influence of various parameters in evaluation model is quantified by golden section search. The field data gathered from the Chi-Chi earthquake of Taiwan in 1999 is also used to perform the verification of evaluation model. The results reveal that the proposed method is simple and effective.
展开▼
