The invention provides a concrete permeability prediction model construction method based on
deep learning, and relates to the technical field of constructional
engineering. The method comprises the following steps: performing
sample preparation and
test design; carrying out a multi-working-condition permeability test;
test data are integrated and cleaned, and modeling preparation
processing is carried out; carrying out
deep learning model training and optimization; and verifying the model precision, and proposing a prediction formula applied to
engineering. Compared with a traditional
algorithm, the optimization efficiency of the method is improved by 3.5 times, the
training period is shortened to be within 30 minutes, the
test set Rreaches 0.999, the equipment cost is reduced by 40% or above, rapid and accurate prediction of the permeability under the complex working condition is achieved, the application threshold is reduced through
engineering expression, the method conforms to the
green engineering and the double-carbon concept, and the method is suitable for large-scale popularization and application. And the anti-seepage design requirements of engineering scenes such as island-
reef hydraulic reclamation and gas storage facilities are met.