Compaction technology has been developed over the past three decades. However, the absence of an effective model that can rapidly and accurately simulate the compaction process limits the further development of compaction technology. In this paper, we propose a vibration compaction process model to simulate the compaction of rockfill materials. A viscoelastic-plastic dynamic compaction deformation formula for rockfill materials was proposed and, considering the horizontal movement of the roller, this formula was utilized for simulating pass-by-pass compaction by dividing the rockfill materials into equal-width blocks. Furthermore, an inversion method based on neural networks was proposed to determine the model parameters of real-world rockfill materials through field test results. Then, the effects of number of roller passes and vibration frequency were analyzed. The experimental results verified that the proposed model could effectively simulate the roller-soil interaction and rockfill material deformation characteristics and could thus provide a theoretical basis for intelligent compaction.

压实技术在过去三年中得到了发展。然而，缺乏能够快速准确地模拟压实过程的有效模型，限制了压实技术的进一步发展。在本文中，我们提出了一种振动压实过程模型来模拟堆石料的压实。提出了堆石料粘弹塑性强夯变形公式，并考虑压路机的水平运动，将堆石料分割成等宽块体，利用该公式模拟路旁压实。此外，提出了一种基于神经网络的反演方法，通过现场试验结果确定现实世界堆石料的模型参数。然后，分析了辊道次数和振动频率的影响。试验结果表明，该模型能够有效模拟碾压土相互作用和堆石料变形特性，为智能压实提供理论依据。