Research on Virtual Simulation Experimental Teaching of Tunnel Subsidence Prediction Based on BP Neural Network
With the rapid development of artificial intelligence technology,traditional civil engineering courses urgently need to be transformed,and virtual simulation experimental teaching is an important part of improving the teaching quality of intelligent construction in universities.With the help of Matlab simulation software,a multilayer BP neural network prediction model is built to predict the ground settlement during the construction of subway shields,and analyze the influence of different input variables and the number of hidden layers on the prediction performance.The results show that for a single hidden layer neural network,it is recommended that if the input variable is 1,then the number of hidden layers should be 1 to 6;for a double hidden layer neural network,it is recommended that if the input variable is 5 then the number of hidden layers in the first and second layers should be 1 to 3.The established single/multi-lay er BP neural network can well predict the trend of settlement changes,and can provide a reference for virtual simulation experimental teaching in intelligent construction courses.
万方数据
维普
