公伯峡混凝土面板堆石坝位移反演分析

采用基于综合应用人工神经网络和演化算法的位移反演分析方法，对在建的公伯峡面板堆石坝坝体沉降变形的现场观测结果进行了反演分析，并通过与坝料室内试验和现场载荷试验结果对比，讨论分析了4种主要坝料在现场条件下的变形特性、分析结果表明，公伯峡面板堆石坝坝料现场的填筑质量较好，主要坝料的变形模量均大于相应由试验室试验得到的模量值．3BⅡ砂砾石料在现场碾压条件下可获得较高的变形模量值，是一种优良的筑坝材料．

Displacement back analysis of Gongboxia Concrete-faced Rockfill Dam

Using in-situ observation data from the CFRD of Gongboxia, the settlement is back analyzed by means of the developed displacements back analysis method, which is based on artificial neural network and evolutionary algorithm. The deformation properties of four main rockfill materials are discussed by comparing the results of back analysis, laboratory tests and field loading tests. It is demonstrated that the deformation modulus of rockfill materials are greater than those from laboratory tests, and the construction of dam is of good quality. The gravel material, 3BII, can achieve high deformation modulus under field roller compaction, which is a suitable construction material for rockfill dam.