No description available.

No description available.

Climate change, uncertainties in energy prices, and the Covid-19 pandemic have significantly reshaped building management, highlighting the need for energy-efficient, safe, and comfortable indoor environments. With advancements in Internet of Things (IoT) sensors and Artificial Intelligence (AI) techniques, optimising building performance now includes forecasting key parameters and intelligently controlling Heating, Ventilation and Air Conditioning (HVAC) systems. However, existing studies often lack practical applicability in real-world scenarios, typically relying on extensive data collection or tailored physical/mathematical models, with limited focus on deployment, scalability, and long-term performance. This thesis addresses the problem from a different angle, proposing an adaptive and practical AI-based solution for energy-efficient comfort optimisation in indoor environments. The designed approach continuously learns from the monitored environment through collected data and requires minimal human effort for configuration and maintenance. The contributions are as follows: i) a method for accurately predicting key parameters using a limited window of data, with a dynamic mechanism to keep the AI model current with environmental changes and operational in a short time frame, and ii) a novel algorithm called EECO for automated and intelligent HVAC control, driven by continuous short-term decisions based on long-term predictions to balance thermal comfort and energy consumption, with no need for preliminary knowledge of the local environment. Evaluation results demonstrate that the proposed approach achieves high prediction accuracy, ensures desired thermal comfort, and reduces the energy footprint by up to approximately 16% in a real-world environment, in addition to potentially saving on operating costs.

An electric power system is a complex network consisting of a large number of power generators and consumers interconnected by transmission and distribution lines. One remarkable thing about the electric grid is that there has to be a continuous balance between the amount of electricity generated and consumed at all times. Maintaining this balance is critical for the stable operation of the grid and this task is achieved in the long term, short term and real-time by operating a three-tier wholesale electricity market consisting of the capacity market, the energy market and the ancillary services market respectively. For a demand resource to participate in the energy and the capacity markets, it needs to be able to reduce the power consumption on-demand, whereas to participate in the ancillary services market, the power consumption of the demand resource needs to be varied continuously following the regulation signal sent by the grid operator. This act of changing the demand to help maintain energy balance is called demand response (DR). The dissertation presents novel algorithms and tools to enable residential buildings to participate as demand resources on such markets to provide DR. Residential sector consumes 37% of the total U.S. electricity consumption and a recent consumer survey showed that 88% of consumers are either eager or supportive of advanced technologies for energy efficiency, including demand response. This indicates that residential sector is a very good target for DR. Two broad solutions for residential DR are presented. The first is a set of efficient algorithms that intelligently controls the customers' heating, ventilating and air conditioning (HVAC) devices for providing DR services to the grid. The second solution is an extensible residential demand response simulation framework that can help evaluate and experiment with different residential demand response algorithms. One of the algorithms presented in this dissertation is to reduce the aggregated demand of a set of HVACs during a DR event while respecting the customers' comfort requirements. The algorithm is shown to be efficient, simple to implement and is proven to be optimal. The second algorithm helps provide the regulation DR while honoring customer comfort requirements. The algorithm is efficient, simple to implement and is shown to perform well in a range of real-world situations. A case study is presented estimating the monetary benefit that can be obtained by implementing the algorithm in a cluster of 100 typical homes and shows promising result. Finally, the dissertation presents the design of a python-based object-oriented residential DR simulation framework which is easy to extend as needed. The framework supports simulation of thermal dynamics of a residential building and supports house hold appliances such as HVAC, water heater, clothes washer/dryer and dish washer. A case study showing the application of the simulation framework for various DR implementation is presented, which shows that the simulation framework performs well and can be a useful tool for future research in residential DR. / PHD / The total power generation and consumption has to always match in the electric grid. When there is a mismatch because the generation is less than the load, the match can be restored either by increasing the generation or by decreasing the load. Often, during system stress conditions, it is cheaper to decrease certain loads than to increase generation, and this method of achieving power balance is called demand response (DR). Residential sector consumes 37% of the total U.S. electricity consumption and is largely unexplored for demand response purpose, so the focus of the dissertation is on providing solutions to enable residential houses to provide demand response services. This dissertation presents two broad solutions. The first is a set of efficient algorithms that intelligently controls the customers’ heating, ventilating and air conditioning (HVAC) devices for providing DR services to the grid while keeping their comfort in mind. The second solution is a simulation software that can help evaluate and experiment with different residential demand response algorithms. The first algorithm is for reducing the collective power consumption of an aggregation of residential HVAC, whereas the second algorithm is for making the collective power follow a signal sent by the grid operators. It is shown that the algorithms presented can intelligently control the HVAC devices such that DR services can be provided to the grid while ensuring that the temperatures of the houses remain within comfortable range. The algorithms can enable demand response service providers to tap into the residential demand response market and earn revenue, while the simulation software can be valuable for future research in this area. The simulation software is simple to use and is designed with extensibility in mind, so adding new features is easy. The software is shown to work well for studying residential building control for demand response purpose and can be a useful tool for future research in residential DR.

<div>Currently, people in North America spend roughly 90% of their time indoors. Therefore, it is important to create comfortable, healthy, and productive indoor environments for the occupants. Unfortunately, our resulting indoor environments are still very poor, especially in multi-occupant rooms. In addition, energy consumption in residential and commercial buildings by HVAC systems and lighting accounts for about 41% of primary energy use in the US. However, the current methods for simulating building energy consumption are often not accurate, and various types of occupant behavior may explain this inaccuracy.</div><div>This study first developed artificial neural network models for predicting thermal comfort and occupant behavior in indoor environments. The models were trained by data on indoor environmental parameters, thermal sensations, and occupant behavior collected in ten offices and ten houses/apartments. The models were able to predict similar acceptable air temperature ranges in offices, from 20.6 °C to 25 °C in winter and from 20.6 °C to 25.6 °C in summer. We also found that the comfortable air temperature in the residences was 1.7 °C lower than that in the offices in winter, and 1.7 °C higher in summer. The reason for this difference may be that the occupants of the houses/apartments were responsible for paying their energy bills. The comfort zone obtained by the ANN model using thermal sensations in the ten offices was narrower than the comfort zone in ASHRAE Standard 55, but that using behaviors was wider.</div><div>Then this study used the EnergyPlus program to simulate the energy consumption of HVAC systems in office buildings. Measured energy data were used to validate the simulated results. When using the collected behavior from the offices, the difference between the simulated results and the measured data was less than 13%. When a behavioral ANN model was implemented in the energy simulation, the simulation performed similarly. However, energy simulation using constant thermostat set point without considering occupant behavior was not accurate. Further simulations demonstrated that adjusting the thermostat set point and the clothing could lead to a 25% variation in energy use in interior offices and 15% in exterior offices. Finally, energy consumption could be reduced by 30% with thermostat setback control and 70% with occupancy control.</div><div>Because of many contextual factors, most previous studies have built data-driven behavior models with limited scalability and generalization capability. This investigation built a policy-based reinforcement learning (RL) model for the behavior of adjusting the thermostat and clothing level. We used Q-learning to train the model and validated with collected data. After training, the model predicted the behavior with R2 from 0.75 to 0.80 in an office building. This study also transferred the behavior knowledge of the RL model to other office buildings with different HVAC control systems. The transfer learning model predicted with R2 from 0.73 to 0.80. Going from office buildings to residential buildings, the transfer learning model also had an R2 over 0.60. Therefore, the RL model combined with transfer learning was able to predict the building occupant behavior accurately with good scalability, and without the need for data collection.<br></div><div><div>Unsuitable thermostat settings lead to energy waste and an undesirable indoor environment, especially in multi-occupant rooms. This study aimed to develop an HVAC control strategy in multi-occupant offices using physiological parameters measured by wristbands. We used an ANN model to predict thermal sensation from air temperature, relative humidity, clothing level, wrist skin temperature, skin relative humidity and heart rate. Next, we developed a control strategy to improve the thermal comfort of all the occupants in the room. The control system was smart and could adjust the thermostat set point automatically in real time. We improved the occupants’ thermal comfort level that over half of the occupants reported feeling neutral, and fewer than 5% still felt uncomfortable. After coupling with occupancy-based control by means of lighting sensors or wristband Bluetooth, the heating and cooling loads were reduced by 90% and 30%, respectively. Therefore, the smart HVAC control system can effectively control the indoor environment for thermal comfort and energy saving.</div><div>As for proposed studies in the future, at first, we will use more advanced sensors to collect more kinds of occupant behavior-related data. We will expand the research on more occupant behavior related to indoor air quality, noise and illuminance level. We can use these data to recognize behavior instead of questionnaire survey now. We will also develop a personalized zonal control system for the multi-occupant office. We can find the number and location of inlet diffusers by using inverse design.</div></div>

I samarbete med Norconsult och kund ska denna rapport undersöka inomhusklimatet i ett museum och resultera i ett förslag på ett kontrollsystem för Heating, Vetilation and Air conditioning (HVAC) som genererar stabil temperatur och relativ luftfuktighet (RF) inom rekommenderade gränsvärden för konstbevaring. Genom litteraturstudier, fältstudier, dataanalys och skapandet av en simulerad modell av utrymmet, ska ett förslag om lämplig kontrollmetod av HVAC framläggas. Vid test av olika kontrollmetoder, tog sig MPC-kontrollen sig bäst ut. Litteraturstudien innefattar flera studier och rapporter som använt sig av MPC-kontroll för liknande projekt med fördelaktiga resultat. Resultaten av dataanalysen visar att under stor del av loggnings-perioden avvek temperatur och/eller RF från gränsvärdena. I vissa utrymmen var uppemot 99% av de loggade värdena utanför gränsvärdet för antingen temperatur eller RF. Av samtliga sensorsamplingar på museet låg 48,53% av dessa utanför de satta rekommendationerna. Generellt var luften i museet för varm och torr till syftet konstförvaltning. Vid jämförelse av de tre utvecklade MPC-kontrollerna, visar alla varianterna till att generera stabila resultat. Det framarbetade förslaget indikerar på ett energieffektivare kontrollsystem gentemot befintligt. Om simuleringsmodellen av utrymmet visar skilja sig från verkliga förhållanden, förändras dynamiken i systemet, och inställningarna för MPC-kontrollen gäller ej längre utan behövs konfigureras om. Projektets primära mål om att generera ett optimerat systemförslag som kontrollerar parametrarna temperatur RF, anses ha uppnåtts - om än endast i simulerade resultat. Vidare arbete innefattar att verifiera modellen mot befintligt system, och därefter inkludera kostnadsfunktioner till MPC-kontrollen. Efter systemimplementering bör faktiska utfall jämföras med simulerade utfall för att utvärdera modellens precision. / In collaboration with Norconsult and the customer, this report will examine the indoor climate in the museum and result in a proposal for a control system for HVAC that generates stable temperature and RH within recommended values ​​for the preservation of art. A proposal for an appropriate control method of the HVAC shall be presented through literature studies, field studies, data analysis, and the creation of a simulated model of the environment. The MPC control performed best during the testing of different control methods. The literature study includes several studies and papers that use MPC control for similar projects with beneficial results. The results of the data analysis show that during a large part of the logging period, the temperature and/or RH deviated from the limit values. In some areas, up to 99% of the logged values ​​were outside recommendations. Of all sensor samplings at the museum, 48.53% of these were registered outside limit values. In general, the air in the museum was too hot and dry for art preservation. When comparing the three developed MPC controllers, all variants show to generate stable results. The elaborated proposal indicates a more energy-efficient control system compared to the existing one. If the simulation model of the environment turns out to differ from real conditions, the dynamics of the system change , and the settings for the MPC control no longer apply and need to be reconfigured. The project's primary goal of generating an optimized system proposal that controls the temperature and relative humidity is considered to have been achieved - albeit only in simulated results. Further work includes verifying the model against the existing system and including cost functions to the MPC control. After system implementation, actual outcomes should be compared with simulated outcomes to evaluate model precision.

Room models, currently used for controller tests, assume the room air to be perfectly mixed. A new room model is developed, assuming non-homogeneous room conditions and distinguishing between different sensor positions. From measurement in real test rooms and detailed CFD simulations, a list of convective phenomena is obtained that has to be considered in the development of a model for a room equipped with different HVAC systems. The zonal modelling approach that divides the room air into several sub-volumes is chosen, since it is able to represent the important convective phenomena imposed on the HVAC system. The convective room model is divided into two parts: a zonal model, representing the air at the occupant zone and a second model, providing the conditions at typical sensor positions. Using this approach, the comfort conditions at the occupant zone can be evaluated as well as the impact of different sensor positions. The model is validated for a test room equipped with different HVAC systems. Sensitivity analysis is carried out on the main parameters of the model. Performance assessment and energy consumption are then compared for different sensor positions in a room equipped with different HVAC systems. The results are also compared with those obtained when a well-mixed model is used. A main conclusion of these tests is, that the differences obtained, when changing the position of the controller's sensor, is a function of the HVAC system and controller type. The differences are generally small in terms of thermal comfort but significant in terms of overall energy consumption. For different HVAC systems the cases are listed, in which the use of a simplified model is not recommended. This PhD has been submitted in accordance to the conditions for attaining both the French and the German degree of a PhD, on a co-national basis, in the frame of a statement of the French government from January 18th, 1994. The research has been carried out in the Automation and Energy Management Group (AGE), Department of Sustainable Development (DDD), at the &amp;quot;Centre Scientifique et Technique du Bâtiment&amp;quot; (CSTB) in Marne la Vallée, France, in collaboration with the &amp;quot;Centre Energétique&amp;quot; (CENERG) at the &amp;quot;Ecole Nationale Supérieure des Mines de Paris&amp;quot; (ENSMP), Paris, France and the Technical University of Dresden (TUD), Germany.

Room models, currently used for controller tests, assume the room air to be perfectly mixed. A new room model is developed, assuming non-homogeneous room conditions and distinguishing between different sensor positions. From measurement in real test rooms and detailed CFD simulations, a list of convective phenomena is obtained that has to be considered in the development of a model for a room equipped with different HVAC systems. The zonal modelling approach that divides the room air into several sub-volumes is chosen, since it is able to represent the important convective phenomena imposed on the HVAC system. The convective room model is divided into two parts: a zonal model, representing the air at the occupant zone and a second model, providing the conditions at typical sensor positions. Using this approach, the comfort conditions at the occupant zone can be evaluated as well as the impact of different sensor positions. The model is validated for a test room equipped with different HVAC systems. Sensitivity analysis is carried out on the main parameters of the model. Performance assessment and energy consumption are then compared for different sensor positions in a room equipped with different HVAC systems. The results are also compared with those obtained when a well-mixed model is used. A main conclusion of these tests is, that the differences obtained, when changing the position of the controller's sensor, is a function of the HVAC system and controller type. The differences are generally small in terms of thermal comfort but significant in terms of overall energy consumption. For different HVAC systems the cases are listed, in which the use of a simplified model is not recommended. This PhD has been submitted in accordance to the conditions for attaining both the French and the German degree of a PhD, on a co-national basis, in the frame of a statement of the French government from January 18th, 1994. The research has been carried out in the Automation and Energy Management Group (AGE), Department of Sustainable Development (DDD), at the &amp;quot;Centre Scientifique et Technique du Bâtiment&amp;quot; (CSTB) in Marne la Vallée, France, in collaboration with the &amp;quot;Centre Energétique&amp;quot; (CENERG) at the &amp;quot;Ecole Nationale Supérieure des Mines de Paris&amp;quot; (ENSMP), Paris, France and the Technical University of Dresden (TUD), Germany.