Data-driven Drill Bit Selection for Bored Piles

In the field of bored pile engineering, the selection of the drill bit is very important. It will directly affect the footage rate during the drilling process, which affects the progress and cost of the entire project. At present, in the field of bored piles, most of the drill bit selection depends on the designer's experience. With the continuous improvement of the degree of informatization in the engineering field, a large amount of information such as formation information, bit parameter information and construction efficiency has been accumulated in the project. In order to save the time of bit selection and improve the progress of the project, this paper proposes a data-driven drill bit selection method for bored piles. First, we extract data from the geological report through NER technology, and model the formation information and drill bit efficiency data separately. And then, we construct a model of drill bit selection based on the neural network, and train the model using the above data. Finally, we select the best bit to assist the designer's decision-making.