Research on Surrounding Rock Classification Method Based on Multi-factor Drilling Information
Based on the scientific and technological demonstration project of the Ministry of Communications'Fengjian Expressway in the Three Gorges Reservoir Area',this study carried out research on the intelligent classification method of surrounding rock based on drilling parameters,and achieved the following main results:(1)A total of 1,000 surrounding rock classification indicators were collected through field tests,and a sample database of intelligent classification data of surrounding rock was established.(2)The prediction model of surrounding rock classification index based on SVR and PSO-BP algorithm is established,and the intelligent classification of surrounding rock based on drilling parameters is trained and realized by using the sample library.(3)The model prediction results show that the deviation of the predicted value of the PSO-BP model on the true value fitting reference line is smaller than that of the SVR model.Especially in terms of the maximum error,the PSO-BP model shows a smaller deviation and shows higher fitting accuracy.Especially when predicting the integrity coefficient,the PSO-BP model has higher prediction accuracy.
measurement while drillingintelligent algorithmsurrounding rock classification
万方数据
维普
