R&D

Statistical learning for the optimization of the calculation time of simulations related to the smoke extraction of buildings

Topic: Numerical simulations of smoke extraction

Fire is a major risk to be taken into account in the design of buildings, but the regulations are not adapted to all situations. For example, in Reunion Island, natural ventilation is favored to improve the thermal comfort of the occupants: in case of fire, it impacts the movement of smoke with the wind.
In addition to the regulations, Fire Safety Engineering (FSE) can be used to justify the performance of a fire safety strategy in a personalized manner. It relies in particular on numerical modeling, whose major drawback is the calculation time. An innovative way to significantly reduce the computation time of fluid mechanics is to approximate the physics with Artificial Intelligence techniques.
Thus, the objective of the thesis is to explore statistical methods in order to accelerate the simulations related to the smoke extraction of buildings in the presence of natural ventilation.

Project dates: 2022 - 2025

Contacts:

Development of a DC microgrid with hybrid energy production and storage for energy autonomy in buildings

Topic: Energy management

Currently, the electricity grid is distributed in alternating current. However, the share of solar energy produced by photovoltaic panels, delivering direct current, is becoming more and more important in the energy mix. At the same time, equipment operating natively in direct current, such as LED lighting, office equipment and electric vehicles, are now part of our daily routine. The combination of these elements generates energy losses since their insertion in our network requires multiple conversions from DC to AC and vice versa. Moreover, few standards exist today to allow a secure installation of DC power in buildings. The objective of this thesis is therefore to realize a 100% DC measurement bench in laboratory in order to quantify the possible energy gains for a DC (Direct Current) electrical distribution. The second step is to develop an EMS (Energy Management System) adapted to this type of electrical architecture to intelligently distribute the energy produced between the equipment according to the available solar resource.

Project dates: 2021-2024

Contacts:

IoT and digital twins for predictive maintenance of equipment

Topic: Predictive Maintenance

Within the framework of a collaboration contract between the laboratories ENERGY-LAB, LCIS and the engineering firm Intégrale Ingénierie, the objective of this project is to develop an instrumentation architecture related to the Internet of Things and the maintenance of equipment. The Internet of Things or IoT, is a field of engineering and research that aims to interconnect through various networks and especially the Internet, the objects that surround us. This development comes to answer the need of the industrial field and more particularly in Reunion Island, which is to be able to predict breakdowns and malfunctions of the process due to the material.
The IoT is accompanied by tools, such as digital twins, which help in decision-making based on data captured by the networks. They allow, among other things, to include predictive maintenance natively in their design. Predictive maintenance is a method that consists in anticipating anomalies based on the actual state of the machines. These data will be submitted as input to algorithms trained to detect the precursors of possible defects.

01-02-2022

Contacts: