Summer stage: Benchmark Development for Smart Buildings Control

Stage title: Benchmark Development for Smart Buildings Control

Place: CentraleSupélec, campus Rennes

Supervisor: Stanislav ARANOVSKIY

General context: Energy is the critical sector affecting the quality of life worldwide. Nowadays, vast concerns of sustainability, climate change, and carbon emission motivate multiple research initiatives on the energy management front. Notably, energy management in buildings is one emerging concept showing significant potential. Globally, buildings produce about 30% of CO2 emissions and consume approximately 40% of the world’s total energy, where almost half of this consumption is due to heating, ventilation, and air conditioning systems.

Modern energy systems in buildings must be reactive; they must adapt to varying operating conditions and satisfy the occupants’ expectations in an energy-efficient way. The idea of a flexible energy management system that reacts to the environment gave rise to the concept of smart buildings that are capable of adapt and efficiently operate in response to changes in external and internal conditions, to make decisions in an uncertain environment, and to learn from experience. The core of a smart building is a sophisticated control system, and Control Theory methods are instrumental in addressing this challenge.

Work description: This stage is a part of the research project “Adaptation and Learning for Smart Buildings” and focuses on the benchmark construction for adaptive energy and climate regulation in a time-varying environment. The goal of the stage is:

    • to study the existing in literature (mathematical) models of buildings (HVAC) and to choose which are suitable for the benchmark,
    • to participate in the collection and aggregation of historical data on the weather conditions and occupants’ behavior,
    • to integrate the collected data into static and dynamical models of buildings.

Desired background: English, basic course on systems and signals (e.g., Automation), Matlab experience.