Skip to Main content Skip to Navigation

Building identification within a connected object ecosystem

Abstract : This thesis is devoted to the problem of the identification of a thermal model of a smart building, whose connected objects alleviate the lack of measurements of the physical quantities of interest. The first algorithm deals with the estimation of the open-loop building system, despite its actual exploitation in closed loop. This algorithm is then modified to account for the uncertainty of the data. We suggest a closedloop estimation of the building system as soon as the indoor temperature is not measured. Then, we return to open-loop approaches. The different algorithms enable respectively to reduce the possible bias contained in a connected outdoor air temperature sensor, to replace the costly solar flux sensor by another connected temperature sensor, and finally to directly use the total load curve, without disaggregation, by making the most of the On/Off signals of the connected objects.
Complete list of metadata

Cited literature [165 references]  Display  Hide  Download
Contributor : ABES STAR :  Contact
Submitted on : Thursday, October 29, 2020 - 2:57:11 PM
Last modification on : Wednesday, June 15, 2022 - 8:46:31 PM
Long-term archiving on: : Saturday, January 30, 2021 - 6:23:35 PM


Version validated by the jury (STAR)


  • HAL Id : tel-02983222, version 1



Tahar Nabil. Building identification within a connected object ecosystem. Signal and Image processing. Télécom ParisTech, 2018. English. ⟨NNT : 2018ENST0001⟩. ⟨tel-02983222⟩



Record views


Files downloads