Tweet Share

• Send by email

  Send by email

  

  Artificial Neural Network use for Pile Settling Prediction in Tunnel Works

  Within the context of pressure of available surface space, tunnel works turned out as solutions for e.g. people and goods transportation. Nevertheless, in densely occupied urban centers, these tunnels can interact with the preexistent buildings in a negative way, inducing foundation settlements, which can be the origin of damage to buildings. That said, a tool is developed to interpolate those settlements given input variables related to the foundation, the soil and the tunnel itself. The obtained value can be used as a decision basis and compared to predefined threshold values. The cited tool is an artificial neural network, backpropagation type. The training and testing data sets are obtained from other authors works regarding soft soils and pile foundations. The network achieved a 90% correlation, what the authors assumed viable to start real life use. The solution developed meets the 4.0 tunneling moment, a part of the fourth industrial revolution, in which data science and information era become part of civil engineering routines.

  Recipient :

  Name ^* :

  E-mail address ^* :

  ^* Required fields

  Send   or  Cancel

• Print