大坝安全监控的偏最小二乘回归分析

Analysis for the partial least-squares regression of dam safety monitoring

针对大坝安全监控模型因子间多重相关性问题,本文引入了偏最小二乘回归法,对渗流资料进行了详细的分析。研究分析表明,偏最小二乘回归法能够有效地克服因子间的多重相关性的影响,建立的模型具有很好的稳定性和解释性,相比一般的最小二乘回归法,更具有推广性。

Aiming at dam safety monitoring of multiple factors related issues, the paper introduced a partial least--square regression; the seepage information is carried out a detailed analysis. Research and analysis show that the method can effectively overcome the multiple factors related to the impact of, the model has a good stability and interpretative, compared to the general least--squares regression, more promotion of. Key words：