A new machine learning method has been developed to accurately predict the compaction density of earth-rock dams. This method uses a portable falling weight deflectometer (PFWD) to measure the time history signal of the dam, which is then analyzed and processed using machine learning algorithms to extract six characteristic parameters. These parameters are then used to predict the wet density and dry density of the dam materials. The method was evaluated using a case study of a dam, and the results showed that the root-mean-square error (RMSE) of the predicted wet density was within 0.041 g/cm3 and that of the predicted dry density was within 0.032 g/cm3. This demonstrates that the new method is an accurate and effective way to rapidly detect and control the construction quality of earth-rock dams.
