Advanced building models for predictive control (ABMPC)

You are here

Project description

The current practice in building operation is based on technology of the past century. Building control is set up as a collection of rules (if x then y) and is therefore reactive by nature. However, with the availability of cheap and wireless monitoring and computational advances, AI-based predictive building control is on the verge of commercial breakthrough. With this project, Sirris will collaborate with DeltaQ to innovate the models that are used in AI for building control. There are three focus points:

  1. extension of the greybox approach (in-house developed by DeltaQ),
  2. development of black-box AI
  3. research of Reinforcement Learning (RL) for building control

This research project will advance the state-of-art in predictive building control and validate the innovation with real-life offline data collected in the commercial projects of DeltaQ.

Project partners


Project details

July 2021 - June 2023

Projet subsidié par la Region de Bruxelles-Capital - Innoviris/project gesubsidieerd door het Brussels Hoofdstedelijk (Project subsidised by the Brussels Capital Region - Innoviris):

Innoviris & region_0.png