Prediction of Tunnel Surface Settlement Based on Particle Swarm Optimization Extreme Learning Machine
An improved extreme learning machine model is proposed to predict surface subsidence to improve the accuracy and speed of surface subsidence prediction.The particle swarm optimization algorithm is introduced to optimize the weights and thresholds of the extreme learning machine to improve its prediction effect.Taking Yanshan Overpass Station of Jinan Rail Transit Line 4 as an example,an empirical analysis of the model was carried out.The improved extreme learning machine was used to predict the surface settlement of the shield tunnel,and was compared with the traditional extreme learning machine model.The MSE,RMSE and MAPE of the extreme learning machine model improved by the particle swarm optimization algorithm are reduced by 22%,28%,and 5.3%,respectively,which verify that the extreme learning machine improved by the particle swarm optimization algorithm has better prediction accuracy and prediction speed.Empirical evidence has verified that the improved extreme learning machine has a comparatively good generalization ability.
NETL
NSTL
万方数据
维普
