Multi-target parameter similar material ratio prediction method based on deep learning

The invention discloses a multi-target parameter similar material proportion prediction method based on deep learning, and particularly relates to the technical field of deep-buried tunnel physical model tests, and the method comprises the following steps: determining the composition of similar materials; selecting a plurality of groups of similar materials for proportioning, and performing an orthogonal mixing test by preparing similar material samples with different proportions; performing a multi-target parameter model test on the sample to obtain required multi-target parameters; dividing the test results obtained in the step 3 into a training set and a test set, and inputting the target parameters in the training set as an input layer and the proportions of different materials in the similar materials as an output layer into the neural network for training; and verifying the trained neural network through test set data, and optimizing neural network parameters by adopting a particle swarm algorithm. Acco