Ph. D. Project
Distributed Fault tolerant control for Building Energy Management system based on Model Predictive control
2018/10/01 - 2021/09/30
Buildings consume approximately 40% of the total energy consumption and contribute towards 30% of the total CO2 emissions. The
Building Energy Management system (BEMS) is a key component of Heating Ventilation and Air Conditioning (HVAC) system. If an
instrument fault occurs it may become impossible for the BEMS to maintain a good compromise between thermal comfort of the
occupants and energy consumption. One of the challenges in the research area in modelling, control and fault diagnosis for BEMS is
due to the fact that buildings are large-scale complex systems composed of sub-systems interconnected in a variety of ways, which may
cause more or less complicated dependencies between local behaviours and external phenomena such as weather conditions.
Moreover, the various building components may have different physical structures that require different scales in time or space for
their representation. Implementation of performance monitoring, Fault Detection and Diagnosis (FDD) algorithms and control laws
require adequate modelling tools adapted to those specific systems. The subject of the thesis concerns development of Fault-Tolerant
Control (FTC) for a large-sized HVAC system. By integrating model-based model predictive control and FDD algorithms, the FTC adapts
the HVAC control laws to a set of subsystem faults and can respect prescribed comfort and energy consumption constraints. The
control strategy that results when information about faults, is used to modify the optimization problem of the model predictive
controller will be referred to as Active Fault Tolerant Model Predictive Control (MPC). The studied MPC law will integrate various
criteria related to the BEMS objectives (comfort, energy saving, ...) and also information related to the faults. Then, the MPC will be
synthesised as Multi-Objectives Fault Tolerant control. The developed algorithms will be implemented and validated on a model of a
multi-zone commercial building equipped with a HVAC system via dedicated simulation tools (Matlab, Energyplus).
distributed control, fault tolerant control, building energy management
Control Identification Diagnosis
Conacyt Scholarship (mexican students only)