Prediction and application of mine roadway surrounding rock deformation based on AdaBoost-GA-ELM-model

摘要

Aiming at the shortcomings of one-sole-model with low accuracy and instability in the deformation prediction for mine roadway surrounding rock, this article comes up with an AdaBoost-GA-ELM model, which combines the ideas of AdaBoost algorithm, genetic algorithm and extreme learning machine, is proposed. The verification of engineering example about trough roof and floor section, 101091004 working surface, Tun-Bao coal mine shows that the AdaBoost-GA-ELM model has almost equal shares in the area of mine roadway surrounding rock deformation, which can bring gratifying prediction results, compared to GA- ELM, GA-BP and gray model, the prediction accuracy of which has a better effect, containing certain value for engineering application.