基于多源监测数据与多目标优化的高堆石坝参数反演分析

Parameter inversion analysis of high rockfill dams based on multisource monitoring data and multi-objective optimization

基于300 m级高堆石坝多材料分区与多源监测布置特点,建立分区目标优化函数,并采用第3代非支配排序多目标遗传算法(non-dominated sorting genetic algorithm-Ⅲ,NSGA-Ⅲ)提出了可充分利用多源监测数据并兼顾整体优化与分区协调的多目标参数反演方法。采用该方法对两河口高堆石坝4个主要材料分区开展了施工期参数反演分析,并与单目标反演参数的计算结果进行对比。结果表明:该方法更能适应300 m级高堆石坝参数反演分析,进一步提高参数反演的精度。

Based on the characteristics of multi-material zoning and multi-source monitoring layout of 300 mlevel high rockfill dam,this paper proposes a multi-objective parameter inversion method by establishing partitioned multi-objective functions and using non-dominated sorting genetic algorithm-Ⅲ(NSGA-Ⅲ),which can make full use of multi-source monitoring data and take into account the overall optimization and coordination of different material zoning of dam.Using this method,parameter inversion analysis is carried out for the four main material partitions of the Lianghekou high rockfill dam during the construction period,and the results are compared with the calculation results of single-objective inversion parameters.The results show that this method is more suitable for parameter inversion analysis of 300 m-level high rockfill dams,and the accuracy of parameter inversion is further improved.