Experimental Validation for Distributed Control of Energy HubsBehrunani, Varsha; Heer, Philipp; Lygeros, John
doi: 10.1088/1742-6596/2600/7/072004pmid: N/A
As future energy systems become more decentralised due to the integration of renewable energy resources and storage technologies, several autonomous energy management and peer-to-peer trading mechanisms have been recently proposed for the operation of energy hub networks based on optimization and game theory. However, most of these strategies have been tested either only in simulated environments or small prosumer units as opposed to larger energy hubs. This simulation reality gap has hindered large-scale implementation and practical application of these method. In this paper, we aim to experimentally validate the performance of a novel multi-horizon distributed model predictive controller for an energy hub network by implementing the controller on a complete network of hubs comprising of a real energy hub inter-faced with multiple virtual hubs. The experiments are done using two different network topologies and the controller shows promising results in both setups.
Model predictive control of heating in a low energy single-family houseNielsen, CM; Larsen, KHK; Thorsteinsson, S; Bendtsen, JD
doi: 10.1088/1742-6596/2600/7/072003pmid: N/A
This paper presents a control scheme that uses a Model Predictive Control (MPC) approach to manage the heating system of a low-energy single-family house. The house is equipped with an air-to-water Heat Pump (HP) and individually controlled hydronic underfloor heating circuits for each of its 11 heating zones. The MPC scheme is designed to maintain individual room comfort levels in each zone, while incorporating weather forecasts and following a heating reference to allow for load shifting for periods with low energy prices and high Photovoltaic (PV) production calculated by an upper level in a hierarchical control scheme. The focus of the design has been the model structure that allows for fast solutions to the MPC optimisation problem, while still capturing the high complexity, non-linear dynamics of the building. The controller is tested on a high-fidelity simulation model of the house, achieving disturbance rejection and system stabilisation. The rapid solving time makes repeated experiments and longer simulations feasible.
Degradation-aware data-enabled predictive control of energy hubsBehrunani, Varsha; Zagorowska, Marta; Hudoba de Badyn, Mathias; Ricca, Francesco; Heer, Philipp; Lygeros, John
doi: 10.1088/1742-6596/2600/7/072006pmid: N/A
Mitigating the energy use in buildings, together with satisfaction of comfort requirements are the main objectives of efficient building control systems. Augmenting building energy systems with batteries can improve the energy use of a building, while posing the challenge of considering battery degradation during control operation. We demonstrate the performance of a data-enabled predictive control (DeePC) approach applied to a single multi-zone building and an energy hub comprising an electric heat pump and a battery. In a comparison with a standard rule-based controller, results demonstrate that the performance of DeePC is superior in terms of satisfaction of comfort constraints without increasing grid power consumption. Moreover, DeePC achieved two-fold decrease in battery degradation over one year, as compared to a rule-based controller.
Towards a novel intelligent and fully interactive IoT framework for residential buildingsErfani Moghaddam, M A; Konstantzos, I
doi: 10.1088/1742-6596/2600/7/072009pmid: N/A
Intelligent operation of buildings plays a crucial role in enhancing energy efficiency, especially in commercial buildings. However, residential buildings suffer from lack of management in energy consumption, contributing to global carbon emissions. Challenges include limited autonomous features, lack of scalability, and costly existing solutions. This paper presents a preliminary study toward the realization of a universal, scalable, and low-cost residential control framework to improve energy performance and reduce the cost of existing building retrofitting approaches. The residential system in its envisioned form may include thermostat, dimming, shading, and window-opening controls. For this study, a preliminary simulation-based investigation of a typical house in multiple climates is performed. To this end, the occupancy profiles data to layout different levels of automation of thermostats and shades have been employed, proposing 3 levels based on a variety of thermostat setpoints and shading. Data from the American Time Use Survey is utilized to inform occupancy and form the integrated automatic control strategy. This approach paves the way for building management systems in homes that integrate IoT-enabled devices for seamless operation, supporting decarbonization and creating interactive and smart buildings. The results showed that from 15% up to 70% reduction in heating and cooling loads are obtainable. This underscores the great potential of implementation of smart control systems in residential buildings.
Comparison of different deep neural networks for system identification of thermal building behaviorGölzhäuser, Simon; Frison, Lilli
doi: 10.1088/1742-6596/2600/7/072008pmid: N/A
Having accurate information available about future thermal building behavior can help to make good decisions in various heating control tasks. However, creating precise mathematical models for many different buildings is a complex and time-consuming task, owing to the heterogeneity of the building stock and the behavior of its occupants. In this paper, we propose a DNN-based system identification approach for predicting the room temperature inside a building based on past information and future weather forecasts. We evaluate various state-of-the-art and custom-built DNN architectures for TSF. Besides prediction performance, storage space and inference speed as measures for the respective model’s complexity are also taken into account. Our main contribution is demonstrating the effectiveness of these models in predicting the room temperature for differently parameterized simulated buildings. By using several distinct buildings for training, validation and testing, we additionally show that these models are capable to generalize in a way such that the room temperature for different buildings can be predicted by a single model, without any changes or adaptions.
Cascaded reinforcement learning based supply temperature controlHuang, C; Seidel, S; Bräunig, J
doi: 10.1088/1742-6596/2600/7/072010pmid: N/A
In this work, a Q-learning based supply temperature control approach for a demonstrator building is proposed. The purpose is to improve the temperature behaviour inside the building and to tackle comfort problems such as overheating and undercooling which cannot be coped with by the standard heating curve. The Q-learning controller considers predicted future weather data as system states. This can be shown to be superior to Q-learning controllers without weather prediction. Furthermore, in order for the controller to capture different thermal effects of different time constants, a cascaded control structure is designed: An inner Q-learning based controller deals with thermal effects of smaller time constants. It is wrapped by an outer slower Q-learning controller which can tackle effects of larger time constants. Therewith, further improvement of comfort can be achieved.
Physics-informed machine learning framework to model buildings from incomplete informationKuo, T; Manikkan, S; Bilionis, I; Liu, X; Karava, P
doi: 10.1088/1742-6596/2600/7/072013pmid: N/A
This paper introduces a physics-informed machine learning framework that leverages statistical methods to seamlessly integrate diverse sources of information, enabling the automated generation of building energy models for specific target buildings. The proposed framework comprises five modules: building survey, building asset database, building information schema, multi-class classification, and physics-based energy model. To illustrate the framework’s effectiveness, we present a case study involving a building with two possible baselines. The results demonstrate that our developed framework successfully generates comprehensive building energy models even when faced with incomplete, effectively capturing baseline scenarios.
Integration of occupant voting systems and smart home platforms for collecting thermal feedback in indoor environmentsCallegaro, Nicola; Albatici, Rossano
doi: 10.1088/1742-6596/2600/7/072012pmid: N/A
The assessment of indoor thermal comfort is increasingly shifting from statistical to personalized models and therefore there is a growing interest in collecting feedback on occupants’ perceptions and preferences. Occupant Voting Systems (OVS) are emerging as a widely used tool in Post Occupancy Evaluations (POE) but the level of occupants’ interaction with these data collection devices, their scientific accuracy, and the integration of feedback data with building management systems, especially in residential buildings, still need to be further explored. This paper presents a study conducted on five dwellings, located in Italy, where smart home switches were used as feedback buttons to collect the thermal sensation of the occupants. These buttons were integrated into an open-source smart home platform, MOQA. The developed system is described in its technical features, highlighting the amount of information collected, the response rate and its interoperability with smart home systems. The results show that OVSs still have limitations in terms of occupant engagement and it is still rather complicated to correlate ratings with environmental variables. However, an easier integration, here described, with smart home systems would partially overcome these problems, turning the OVS into a useful tool for both users and research purposes.
Comparison of two deep reinforcement learning algorithms towards an optimal policy for smart building thermal controlSilvestri, Alberto; Coraci, Davide; Wu, Duan; Borkowski, Esther; Schlueter, Arno
doi: 10.1088/1742-6596/2600/7/072011pmid: N/A
Heating, Ventilation, and Air Conditioning (HVAC) systems are the main providers of occupant comfort, and at the same time, they represent a significant source of energy consumption. Improving their efficiency is essential for reducing the environmental impact of buildings. However, traditional rule-based and model-based strategies are often inefficient in real-world applications due to the complex building thermal dynamics and the influence of heterogeneous disturbances, such as unpredictable occupant behavior. In order to address this issue, the performance of two state-of-the-art model-free Deep Reinforcement Learning (DRL) algorithms, Proximal Policy Optimization (PPO) and Soft Actor-Critic (SAC), has been compared when the percentage valve opening is managed in a thermally activated building system, modeled in a simulated environment from data collected in an existing office building in Switzerland. Results show that PPO reduced energy costs by 18% and decreased temperature violations by 33%, while SAC achieved a 14% reduction in energy costs and 64% fewer temperature violations compared to the onsite Rule-Based Controller (RBC).
Comparison of different coupling variants for building and HVAC simulationHuang, C; Eckstädt, E
doi: 10.1088/1742-6596/2600/7/072005pmid: N/A
Coupled building and HVAC simulations applying the FMI standard promise many advantages in terms of system design and optimisation. In this paper, a systematic analysis of possible options for such couplings is presented. For one of these options, a case study is carried out where a quantitative comparison between the uncoupled reference system and the corresponding coupled system with different communication step sizes is carried out. Based on the simulation results, recommendations and challenges for such couplings are derived. In particular, the correlation between the communication step size of the FMUs and the computing time, as well as the deviations of the simulation results compared to the results of an uncoupled simulation are analysed.