Research and Application of a New Prediction Model for Hydraulic Blasting in Tunnel Excavation
To comprehensively and scientifically analyze the intrinsic relationship between multi-dimensional blasting characteristic parameters and blasting effects,machine learning technology is used to deeply mine and analyze the measured data of tunnel excavation.Based on the development and technology of new hydraulic blasting in tunnel excavation,a BO-LightGBM prediction optimization model was proposed to comprehensively predict the blasting effect indicators.And the prediction model was successfully applied in practical engineering projects.The analysis of the application results show that the BO-LightGBM prediction optimization model has certain feasibility,with an average absolute error(MAE)of 3.729,a root mean square error(RMSE)of 5.576,and a coefficient of determination(R2)of 0.783.This provides a new and efficient way for on-site construction technicians to design tunnel blasting excavation,and saves time,labor,and costs for quickly entering the actual blasting process.The prediction model of this new water pressure blasting technology has been identified as internationally leading by the China Blasting Industry Association.
tunnel excavationnew water pressure blastingprediction modelblasting effectmachine learning
NETL
NSTL
万方数据
