Multiple linear regression analysis model and artificial neural network model to calculate and estimate the blast induced area of the tunnel face. A case study Deo Ca tunnel

Abstract

The area of the tunnel face after the blasting is a very important factor in underground excavations where the drilling and blasting method is used. The area of the tunnel face, this is a significant factor that has affected the cost and safety of underground constructions in case of using the drilling and blasting method in underground excavations. Because the area of the tunnel after the blasting depends on many different parameters, such as geological conditions in the area where the tunnel is located, the parameters of the explosion, and other parameters of the tunnel, it is very difficult to accurately determine the value of the tunnel face area after blasting. This paper uses the data obtained in the actual blasting of the Deo Ca tunnel (39 datasets) to build the computational and prediction models for the area of the tunnel face after blasting by two methods, the multiple linear regression analysis method and the method of using artificial neural network (ANN). Determination coefficient R2 of multiple linear regression analysis (MLRA) method and ANN method were obtained at 0.9224, and 0.9449, respectively. The applicability of the multiple linear regression analysis method and ANN method in calculating and predicting tunnel face area after blasting were validated based on a comparison with the results of the tunnel face area after blasting in practice.