scholarly journals Online Implementation of a Soft Actor-Critic Agent to Enhance Indoor Temperature Control and Energy Efficiency in Buildings

Energies ◽  
2021 ◽  
Vol 14 (4) ◽  
pp. 997
Author(s):  
Davide Coraci ◽  
Silvio Brandi ◽  
Marco Savino Piscitelli ◽  
Alfonso Capozzoli

Recently, a growing interest has been observed in HVAC control systems based on Artificial Intelligence, to improve comfort conditions while avoiding unnecessary energy consumption. In this work, a model-free algorithm belonging to the Deep Reinforcement Learning (DRL) class, Soft Actor-Critic, was implemented to control the supply water temperature to radiant terminal units of a heating system serving an office building. The controller was trained online, and a preliminary sensitivity analysis on hyperparameters was performed to assess their influence on the agent performance. The DRL agent with the best performance was compared to a rule-based controller assumed as a baseline during a three-month heating season. The DRL controller outperformed the baseline after two weeks of deployment, with an overall performance improvement related to control of indoor temperature conditions. Moreover, the adaptability of the DRL agent was tested for various control scenarios, simulating changes of external weather conditions, indoor temperature setpoint, building envelope features and occupancy patterns. The agent dynamically deployed, despite a slight increase in energy consumption, led to an improvement of indoor temperature control, reducing the cumulative sum of temperature violations on average for all scenarios by 75% and 48% compared to the baseline and statically deployed agent respectively.

2015 ◽  
Vol 108 ◽  
pp. 205-214 ◽  
Author(s):  
Rui M. Novais ◽  
G. Ascensão ◽  
M.P. Seabra ◽  
J.A. Labrincha

2011 ◽  
Vol 43 (11) ◽  
pp. 3019-3026 ◽  
Author(s):  
Roberta Ansuini ◽  
Roberto Larghetti ◽  
Alberto Giretti ◽  
Massimo Lemma

2011 ◽  
Vol 361-363 ◽  
pp. 1047-1050
Author(s):  
Bin Liao

The pattern of using the household billing to promote heating energy savings has become a focus discussion in the current national energy conservation. Nowadays the average energy consumption in China is 2 to 3 times than the developed countries with the same weather conditions, equivalent to the level of developed countries in 60 to 70 years. We report a daily heat-energy consumption measuring test in Beijing since 2009, the result shows that 90% of the total households we tested never change their valves to regulate the heat exchange systems in two winters, the one at least change their valves once are about 5%. So that the way households use the central heating is not fit for the need to save heat-energy.


Author(s):  
Lanbin Liu ◽  
Lin Fu ◽  
Yi Jiang

A large-scale survey and on-site measurements on space heating systems in Beijing has been carried out since 2005. Detailed analysis shows that the improvement of system regulation to adjust the heating demand and to avoid over-heating in building space is the key to reduce the heating energy consumption. It also indicates that combined heat and power (CHP) based district heating network is the most suitable solution for the space heating in Chinese northern cities. Thus, the priority should be in the research and development of new techniques to improve heating system regulation and control. In China, there are three reasons for poor heating system regulation: • the lack of control devices in space heating system, • the complex and inconvenient operation, and • the insufficient motivation because the charging policy is based on the heating areas. Field test results show that 20% to 30% of thermal energy is wasted because of the poor heating system regulation. In order to solve these problems, a novel “wireless on-off control” system for household heat adjusting and metering has been proposed. This technology works in the following way: 1) a calorimeter is installed at each building to measure the total heat consumption of the whole building; 2) an on-off valve is installed for each household and an indoor temperature controller is provided for the occupants. The operation procedure is as follows. First, the desired indoor temperature is set by the users through the indoor temperature controller and wireless signals are sent to the on-off valve. Then the on-off valve detects the real indoor temperature and determines the difference between the real temperature and the set value. After this, the valve calculates the proportion of on-time to off-time in the next step according to the thermal strategies programmed in the valve’ CPU. Then the valve is controlled according to the proportion to maintain the desired indoor temperature; and 3) the heating cost of each household can be allocated according to its heating area and the accumulative open time of the valve. The proposed technique has been applied in fifteen residential communities with the total areas of 1,200,000 m2. The testing results show that: 1) indoor temperatures were accurately controlled within +/− 0.5 °C around the set point, so that the problem of overheating can be avoided; 2) the difference of temperature in different room is less than 1 °C. Therefore, if radiators in each room are designed reasonably, to control the temperature of one room can meet the requirements of the entire user’s apartment; 3) energy consumption in the controlled household was approximately 30% lower than the uncontrolled household with the same building type.


Sign in / Sign up

Export Citation Format

Share Document