A hybrid monitoring model of rockfill dams considering the spatial variability of rockfill materials and a method for determining the monitoring indexes

Abstract

To further enhance the prediction accuracy of the monitoring model and improve the rationality of the method for determining the monitoring indexes, this paper considers the impact of the spatial variability of rockfill materials on the displacement of the dam. Based on the stochastic finite element method (SFEM), a new rockfill dam displacement monitoring hybrid model is established, and a method to determine the hybrid index for rockfill dam displacement monitoring is proposed by multiindex amalgamation. First, the SFEM is used to simulate the spatial variability of the rockfill mechanical parameters in the process of calculating the water pressure component and the time-dependent component. On this basis, the expression of each component is fitted to construct a stochastic finite element method hybrid model (SFEMH model) of the rockfill dam. Subsequently, by predicting the displacement of the dam and separating each displacement component, the displacement monitoring hybrid index of the rockfill dam is determined by combining the confidence interval method and the most unfavourable components. Finally, the SFEMH model and hybrid index are applied to analyse the actual monitoring data of a rockfill dam. The results show that the proposed model and the method for determining the displacement monitoring hybrid index are scientific and reasonable. The prediction accuracy and effect of the proposed novel model outperform those of other methods in terms of many evaluation indicators, and the reliability of the monitoring indexes is significantly improved. The proposed hybrid model and hybrid index provide a new method for the efficient operation management and accurate safety performance evaluation of rockfill dams.