Balancing Energy Consumption and Thermal Comfort with Deep Reinforcement Learning

Author(s):  
Franco Cicirelli ◽  
Antonio Guerrieri ◽  
Carlo Mastroianni ◽  
Luigi Scarcello ◽  
Giandomenico Spezzano ◽  
...  
Energies ◽  
2020 ◽  
Vol 13 (23) ◽  
pp. 6354
Author(s):  
Yassine Chemingui ◽  
Adel Gastli ◽  
Omar Ellabban

Energy efficiency is a key to reduced carbon footprint, savings on energy bills, and sustainability for future generations. For instance, in hot climate countries such as Qatar, buildings are high energy consumers due to air conditioning that resulted from high temperatures and humidity. Optimizing the building energy management system will reduce unnecessary energy consumptions, improve indoor environmental conditions, maximize building occupant’s comfort, and limit building greenhouse gas emissions. However, lowering energy consumption cannot be done despite the occupants’ comfort. Solutions must take into account these tradeoffs. Conventional Building Energy Management methods suffer from a high dimensional and complex control environment. In recent years, the Deep Reinforcement Learning algorithm, applying neural networks for function approximation, shows promising results in handling such complex problems. In this work, a Deep Reinforcement Learning agent is proposed for controlling and optimizing a school building’s energy consumption. It is designed to search for optimal policies to minimize energy consumption, maintain thermal comfort, and reduce indoor contaminant levels in a challenging 21-zone environment. First, the agent is trained with the baseline in a supervised learning framework. After cloning the baseline strategy, the agent learns with proximal policy optimization in an actor-critic framework. The performance is evaluated on a school model simulated environment considering thermal comfort, CO2 levels, and energy consumption. The proposed methodology can achieve a 21% reduction in energy consumption, a 44% better thermal comfort, and healthier CO2 concentrations over a one-year simulation, with reduced training time thanks to the integration of the behavior cloning learning technique.


2019 ◽  
pp. 53-65
Author(s):  
Renata Domingos ◽  
Emeli Guarda ◽  
Elaise Gabriel ◽  
João Sanches

In the last decades, many studies have shown ample evidence that the existence of trees and vegetation around buildings can contribute to reduce the demand for energy by cooling and heating. The use of green areas in the urban environment as an effective strategy in reducing the cooling load of buildings has attracted much attention, though there is a lack of quantitative actions to apply the general idea to a specific building or location. Due to the large-scale construction of high buildings, large amounts of solar radiation are reflected and stored in the canyons of the streets. This causes higher air temperature and surface temperature in city areas compared to the rural environment and, consequently, deteriorates the urban heat island effect. The constant high temperatures lead to more air conditioning demand time, which results in a significant increase in building energy consumption. In general, the shade of the trees reduces the building energy demand for air conditioning, reducing solar radiation on the walls and roofs. The increase of urban green spaces has been extensively accepted as effective in mitigating the effects of heat island and reducing energy use in buildings. However, by influencing temperatures, especially extreme, it is likely that trees also affect human health, an important economic variable of interest. Since human behavior has a major influence on maintaining environmental quality, today's urban problems such as air and water pollution, floods, excessive noise, cause serious damage to the physical and mental health of the population. By minimizing these problems, vegetation (especially trees) is generally known to provide a range of ecosystem services such as rainwater reduction, air pollution mitigation, noise reduction, etc. This study focuses on the functions of temperature regulation, improvement of external thermal comfort and cooling energy reduction, so it aims to evaluate the influence of trees on the energy consumption of a house in the mid-western Brazil, located at latitude 15 ° S, in the center of South America. The methodology adopted was computer simulation, analyzing two scenarios that deal with issues such as the influence of vegetation and tree shade on the energy consumption of a building. In this way, the methodological procedures were divided into three stages: climatic contextualization of the study region; definition of a basic dwelling, of the thermophysical properties; computational simulation for quantification of energy consumption for the four facade orientations. The results show that the façades orientated to north, east and south, without the insertion of arboreal shading, obtained higher values of annual energy consumption. With the adoption of shading, the facades obtained a consumption reduction of around 7,4%. It is concluded that shading vegetation can bring significant climatic contribution to the interior of built environments and, consequently, reduction in energy consumption, promoting improvements in the thermal comfort conditions of users.


Author(s):  
Jun Long ◽  
Yueyi Luo ◽  
Xiaoyu Zhu ◽  
Entao Luo ◽  
Mingfeng Huang

AbstractWith the developing of Internet of Things (IoT) and mobile edge computing (MEC), more and more sensing devices are widely deployed in the smart city. These sensing devices generate various kinds of tasks, which need to be sent to cloud to process. Usually, the sensing devices do not equip with wireless modules, because it is neither economical nor energy saving. Thus, it is a challenging problem to find a way to offload tasks for sensing devices. However, many vehicles are moving around the city, which can communicate with sensing devices in an effective and low-cost way. In this paper, we propose a computation offloading scheme through mobile vehicles in IoT-edge-cloud network. The sensing devices generate tasks and transmit the tasks to vehicles, then the vehicles decide to compute the tasks in the local vehicle, MEC server or cloud center. The computation offloading decision is made based on the utility function of the energy consumption and transmission delay, and the deep reinforcement learning technique is adopted to make decisions. Our proposed method can make full use of the existing infrastructures to implement the task offloading of sensing devices, the experimental results show that our proposed solution can achieve the maximum reward and decrease delay.


2014 ◽  
Vol 493 ◽  
pp. 74-79
Author(s):  
Y.A. Sabtalistia ◽  
S.N.N. Ekasiwi ◽  
B. Iskandriawan

Energy consumption for air conditioning systems (air conditioning system) increased along with the increasing need for fresh air and comfortable in the room especially apartments. FAC system (Floor Air Conditioning) is growing because it is more energy efficient than CAC (Ceiling Air Conditioning) system. However, the position of the AC supply is on the lower level at the FAC system causes draft discomfort becomes greater as air supply closer to the occupants so that thermal comfort can be reduced. Heat mixture of windows, exterior walls, kitchen, and occupants in the studio apartment affect thermal comfort in the room too.This study aims to determine the position of the AC supply which has the best thermal comfort of FAC system in the studio apartment. It can be done by analyzing ADPI (Air Diffusion Performance Index), the distribution of air temperature, wind speed, RH (Relative Humidity), and DR (Draft Risk) to change the position of the AC supply supported by CFD (Computational Fluid Dynamics) simulation.This result prove that AC position 2 (on wall near the kitchen) is more comfortable than AC position 1 (on the bathroom wall) because AC position 2 away from occupied areas, thereby reducing the occurrence of draught discomfort.


2020 ◽  
pp. 014459872096921
Author(s):  
Yanru Li ◽  
Enshen Long ◽  
Lili Zhang ◽  
Xiangyu Dong ◽  
Suo Wang

In the Yangtze River zone of China, the heating operation in buildings is mainly part-time and part-space, which could affect the indoor thermal comfort while making the thermal process of building envelope different. This paper proposed to integrate phase change material (PCM) to building walls to increase the indoor thermal comfort and attenuate the temperature fluctuations during intermittent heating. The aim of this study is to investigate the influence of this kind of composite phase change wall (composite-PCW) on the indoor thermal environment and energy consumption of intermittent heating, and further develop an optimization strategy of intermittent heating operation by using EnergyPlus simulation. Results show that the indoor air temperature of the building with the composite-PCW was 2–3°C higher than the building with the reference wall (normal foamed concrete wall) during the heating-off process. Moreover, the indoor air temperature was higher than 18°C and the mean radiation temperature was above 20°C in the first 1 h after stopping heating. Under the optimized operation condition of turning off the heating device 1 h in advance, the heat release process of the composite-PCW to the indoor environment could maintain the indoor thermal environment within the comfortable range effectively. The composite-PCW could decrease 4.74% of the yearly heating energy consumption compared with the reference wall. The optimization described can provide useful information and guidance for the energy saving of intermittently heated buildings.


2021 ◽  
Vol 8 ◽  
Author(s):  
Huan Zhao ◽  
Junhua Zhao ◽  
Ting Shu ◽  
Zibin Pan

Buildings account for a large proportion of the total energy consumption in many countries and almost half of the energy consumption is caused by the Heating, Ventilation, and air-conditioning (HVAC) systems. The model predictive control of HVAC is a complex task due to the dynamic property of the system and environment, such as temperature and electricity price. Deep reinforcement learning (DRL) is a model-free method that utilizes the “trial and error” mechanism to learn the optimal policy. However, the learning efficiency and learning cost are the main obstacles of the DRL method to practice. To overcome this problem, the hybrid-model-based DRL method is proposed for the HVAC control problem. Firstly, a specific MDPs is defined by considering the energy cost, temperature violation, and action violation. Then the hybrid-model-based DRL method is proposed, which utilizes both the knowledge-driven model and the data-driven model during the whole learning process. Finally, the protection mechanism and adjusting reward methods are used to further reduce the learning cost. The proposed method is tested in a simulation environment using the Australian Energy Market Operator (AEMO) electricity price data and New South Wales temperature data. Simulation results show that 1) the DRL method can reduce the energy cost while maintaining the temperature satisfactory compared to the short term MPC method; 2) the proposed method improves the learning efficiency and reduces the learning cost during the learning process compared to the model-free method.


Author(s):  
Danial Mohammadi ◽  
Simin Nasrabadi

Background: One way to achieve a standard heating, ventilating, and air conditioning system with maximum satisfaction is to use a thermal index to identify and determine the thermal comfort of people. In this study we intend to evaluate thermal comfort based on PMV-PPD (Predicted Mean Vote/Predicted Percentage Dissatisfied) model in workers of screening center for COVID-19. Methods: The study period was from March 1 to October 31, 2020. In this study, we used the ISO 7730 model to determinate PMV-PPD index. PMV index was used to determine thermal comfort at different scales in Birjand city with arid and hot climate. All data were analyzed using R software (version 3.3.0) and IBM SPSS statistics softwares. Results: The maximum and minimum recorded physical PMV values in the study period were observed in June as (2.09 ± 0.03) and March as (-1.27 ± 0.14), respectively. The amplitude of the thermal sense in the study period was varied between slightly cool (-1.5) and warm (+2.5). The PPD in spring was 40% which indicated slightly warm to hot condition. Conclusions: The October was the only month during the study in which thermal stress was in comfort or neutral thermal condition.  Our results suggest that thermal comfort has dimensions and indices which are helpful in managing energy consumption.


Sign in / Sign up

Export Citation Format

Share Document