Accounting for uncertainties in dam safety assessment via machine learning techniques

When working with massive and complex structures such as dam-type structures it is required to generate a scale numerical model to simulate the real behaviour of the dam. The problem of this approach is that to be able to generate all possible hazard scenarios and system configuration, several computing hours are required. This represents a huge problem when the decision-making process need to be done in presence of natural disaster such as earthquakes. Thus, the main goal of this research is to replace the numerical model with a mathematical algorithm, which is orders of magnitude faster, to be able to predict the response of the dam under a seismic event. This research will allow Hydro-Quebec to implement this new methodology to provide quick answers to the owners and the society for a safety assessment.

Faculty Supervisor:

Patrick Paultre

Student:

Rocio Segura

Partner:

Discipline:

Engineering - civil

Sector:

Energy

University:

Program:

Accelerate