Airflow pattern control using artificial intelligence for effective removal of indoor airborne hazardous materials. (15th October 2021)

• Record Type: Journal Article
• Authors: Kim, Na Kyong; Kang, Dong Hee; Lee, Wonoh; Kang, Hyun Wook
• Is Part Of: Building and environment. Volume 204(2021)
• Publication Date: 2021-10-15
• Subjects: Ventilation control -- Artificial intelligence (AI) -- Particulate matter (PM) -- Indoor air quality (IAQ) -- Airflow pattern -- Airborne hazardous material; Buildings -- Environmental engineering -- Periodicals; Building -- Research -- Periodicals; Constructions -- Technique de l'environnement -- Périodiques; Electronic journals; 696
• DOI: 10.1016/j.buildenv.2021.108148
• Languages: English
• ISSNs: 0360-1323

Abstract

Abstract: In the field of environment and building engineering, the airflow pattern is one of the main factors controlling the indoor environmental quality. The control of airflow patterns has been examined, yet it remains problematic owing to a number of factors related to the location of the air inlets and outlets, and distributions of the indoor pollutants. Herein, we present a novel ventilation control strategy using artificial intelligence to rapidly remove hazardous airborne materials in an isolation room. The isolation room was designed with nine inlets on the ceiling and six outlets at the floor level to control airflow patterns through selective on/off switching of the inlets. To build a database of the indoor environment, numerical simulations were performed on different distributions of airborne particles and airflow patterns. We preprocessed the dispersion data of the airborne particles by discretizing the simulated volume into several cuboids and averaging the concentration data for use as the input variables for artificial intelligence. The artificial intelligence model predicted an efficient ventilation condition based on the distribution of the airborne materials within a prediction accuracy of 91%. The controlled strategy decreased the removal time up to maximum 63.65%, compared to conventional ventilation system. Furthermore, the proposed control strategies for airflow patterns can effectively prevent the spread of infectious viruses and reduce the risk of indoor infection transmission. Highlights: Novel strategy of airflow patterns to control the indoor air quality. Artificial neural network successfully predicting airborne hazardous materials removing. Efficient airborne hazardous materials removing via indoor airflow pattern control. Artificial intelligence technology applied to indoor air quality and energy consumption.