Exploiting discrete structures in optimizing a dynamic air quality monitoring system
Báo cáo viên: Nguyễn Phi Lê

Thời gian: 9h30, Thứ 5, ngày 5//03/2020.
Địa điểm: Phòng 612, nhà A6, Viện Toán học, 18 Hoàng Quốc Việt.

Tóm tắt:Recently, with rapid industrialization and urbanization, air pollution is becoming an increasingly painful issue than ever in Vietnam. As one of the biggest cities in the country, Hanoi has suffered from heavy air pollution. Since the end of 2019, we have been witnessing prolonged air pollution with constantly heavy smog and the AQI indicator at an alarming level. In such a situation, a thorough solution for monitoring air quality on a large scale and predicting future air quality is essential to help people protecting their health from air pollution, as well as the government planting policies in time.

So far, air monitoring has been carried out by using monitoring stations located at fixed locations. However, due to the cost of installation, deployment, and operation, the number of monitoring stations deployed is often very small. As a result, they only cover a limited area, which is insufficient compared to the actual needs.

In this research, we propose a novel air monitoring system that exploits the dynamicity of public vehicles to broaden the air quality monitoring regions. We exploit state-of-the-art optimization techniques to design the system that can minimize the costs and optimize the performance. We also leverage novel deep learning models to accurately forecasting the air quality in the future.

In this talk, we first present the overview of the proposed system and figure out the research problems inside. Then, we focused on the problem of optimizing the monitoring positions and show how to exploit discrete structures in determining the optimal solution.

Trở lại

Công bố khoa học mới