Road water-stable base cracking performance prediction method based on neural network learning

The invention discloses a road water-stable base cracking performance prediction method based on neural network learning, and the method comprises the following steps: firstly, carrying out the CT scanning of a cement-based test piece, carrying out the segmentation and three-dimensional reconstruction of an image through software, and calculating the orientation distribution index and void ratio of each particle size of the test piece; secondly, the model is imported into ABAQUS for a uniaxial tension and compression numerical test, and the compression modulus and the tensile modulus of the model are obtained; thirdly, constructing a neural network model, and training by taking the orientation distribution index and void ratio of each particle size of the test piece as input data and taking the compression modulus and the tensile modulus as output data; and finally, performing core drilling and sampling on the cracked road section of the semi-rigid base, performing scanning reconstruction, calculating the mat