Continuous Identification of Soil Stiffness Considering the Nonlinear Behavior of Vibratory Roller-Soil System

Continuous monitoring of soil compaction state by an instrumented roller requires the model that can capture the nonlinear behavior during roller-soil vibration, the establishments of roller-soil dynamic models and indices that can accurately reflect the soil compaction state are two important components of RICM technology. This paper proposed a roller-soil lumped model considering the nonlinear behavior and use the harmonic balance identification method to identify the soil stiffness and the nonlinear terms of roller-soil model. First, fast Fourier transform was used to perform the frequency spectrum of roller acceleration signals, then, an equation was established and solved by the least square method to calculate the nonlinear terms of roller-soil model and soil stiffness. Based on the simulation experiment, the test results were seen to accurately predict the real values of a nonlinear roller-soil lumped model and the error of estimation was always less than 5% and decreased with an increase in the harmonic numbers. Finally, soil stiffness was proposed as a compaction quality assessment index.