Back to search

Technical University of Crete (TUC)

Model-assisted control for energy efficiency in buildings

Abstract

dc:description

Effective utilization of energy in buildings is receiving significant attention. This interest is justified on the observation that buildings account for a significant portion of end-energy use. In Europe 40% of the total energy consumed is used for the operation of buildings with a significant part of that energy used for conditioning occupied spaces. Thus the value of effective energy utilization during the building operational phase is undisputed both in terms of achieving good occupant comfort and in reducing energy consumption.In energy management of buildings, Building Management Systems (BMS) (or Building Automation and Control Systems (BACS)) play an essential role in managing buildings by providing some monitoring and control capability for multiple sub-systems including the Heating Ventilation and Air Conditioning (HVAC) system. Very often, the installation and commissioning of a BMS system is viewed largely as a quality control process, and more specifically the target is proper installation and ensuring a minimum-level of functionality, rather than good operational performance. This has enabled the utilization of flexible and scalable BMS solutions. Here, the BMS is divided into two parts: Local and Remote. The local part is responsible for managing local control loops (usually PID), data collection and transfer, and very simple analytics; and the remote layer is responsible for providing more advanced services. The availability of such a cloud-based platform enables the development and utilization of advanced building energy management services, which cannot be deployed in local hardware due to their computational requirements and allow for semi-automatic intelligent control.One of the most recent and widely-applied methodologies for designing intelligent building control strategies is Model Predictive Control (MPC). Here, a model of the building is available and using available weather and occupancy forecasts an optimization process determines a set of control actions to be applied to the building for a pre-defined period of time (e.g. 24 hours). The different control strategies are evaluated in the optimizer on the basis of a cost function and a set of (visual, thermal, etc.) comfort constraints. Then, the first control command is applied to the building and the whole process repeats. This approach has certain limitations, mainly regarding the inability to incorporate in a structured manner more elaborate thermal comfort indices, such as the Fanger index, and the requirement for linear or bi-linear building models, in order to define a convex optimization task. The methodology developed in this thesis tries to address these limitations while maintaining the benefits of the MPC paradigm. We have developed a framework to deliver optimized and integrated operation of all energy-influencing components of a building that include generation, distribution and emission elements. First, a set of parametric control functions (controllers) are identified for each emission/distribution/generation system, following guidelines available in international standards. Next, given (weather, occupancy, etc.) forecasts for a predefined time window, say three days, a surrogate-based stochastic optimization algorithm is used to create candidate controller parameters to be applied to the building, and a detailed thermal simulation model (designed using detailed thermal simulation software like e.g. EnergyPlus or TRNSYS) is used to evaluate these candidate solutions. The evaluation is performed on the basis of a defined cost function and a set of (visual, thermal, air-quality, etc.) comfort constraints. The final, improved controllers are communicated to the building-side and are applied to the building.In order to verify the efficiency of the proposed methodology, a hierarchy of experiments has been designed, facilitating simulation-based studies and real building experiments in two study buildings located in Greece and Germany. The simulation-based studies are used to evaluate the potential of the proposed approach in a controlled and disturbance-free environment, while the goal of the real building experiments is to study the behavior of the control design process under real-world conditions, influenced by user- and weather-induced disturbances.In all our experiments the proposed methodology was able to outperform the baseline control strategy of the buildings (usually hard-coded in a set of pre-defined rules) in terms of total energy consumed, while preserving comfortable interiors. This is because a static, rule-based controller is unable to properly account for all the predicted and un-predicted disturbances (e.g. weather conditions, occupancy, setpoint tracking errors, etc.), whereas the model-based control design approach, utilizing the thermal simulation model of the building and the automated control design setup, was able to make informative decisions and control the building in an efficient manner. An important aspect investigated through the experiments is the topic of (thermal) comfort-based control, since it was clear from the results that a discussion on energy efficiency seems untimely when discussed without any reference to comfort as there is a clear trade-off between comfort and energy. Having methods that allow the automated selection of building operation to desired levels of comfort is both desirable and can have significant implications. On the other hand, in existing state of the art, comfort is often synonymous with the temperature in the building zones being maintained within certain (prescribed) limits. As we investigate in the thesis this assumption while true in certain cases, in many cases is not sufficient to ensure true comfort for the building occupants, with detrimental effects to health and productivity (when office buildings are considered). In the proposed approach, we define comfort based on ISO 7730, which supports the definition and use of the Fanger index which captures, among other parameters, the influence of air and radiant temperatures along with humidity.

Degree

thesis:*
Grantor dc:publisher
Technical University of Crete (TUC)
Year dc:date
2017

Author and committee

dc:creator, dc:contributor.*
Authors dc:creator
  • Kontes, Georgios
  • Κόντες, Γεώργιος

Subjects

dc:subject × 10

Rights

Language dc:language
eng

Identifiers

dc:identifier.*
Identifier
10.12681/eadd/41560

Chain of custody

source
Harvested from
Greek National Archive of PhD Theses
Base URL
phdtheses.ekt.gr/eadd_oai/request
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
citation

Kontes, Georgios; Κόντες, Γεώργιος. Model-assisted control for energy efficiency in buildings. Technical University of Crete (TUC), 2017. http://hdl.handle.net/10442/hedi/41560