基于主成分的大坝监测资料时变预测模型

Time-varying Prediction Model of Dam Monitoring Data Based on Principal Component Analysis

由于内外各种因素的影响,大坝安全监控参数会随时间而变化,而常规监控模型常常采用非时变的参数。基于主成分分析,利用缩减后的主成分荷载建立了时变预测模型。实例表明,该模型可以减少计算时间,有效削弱因子多重相关性的影响,提高大坝监测效应量的预测水平。

As influenced by many internal and external factors, the dam safety monitoring parameters will change over time, and the time-invariant parameters arc often used in conventional monitoring models. Based on principal component analysis, a time-varying prediction model is set up by using reduced principal component loads. The case studies show that the model can save computational time, weaken muhi-correlativity among factors and improve the effect size prediction of dam monitoring.