Related Experiment Video
Updated: May 13, 2025
08:59
Measuring Sub-23 Nanometer Real Driving Particle Number Emissions Using the Portable DownToTen Sampling System
Published on: May 22, 2020
5.4K
An optimized TCN-LSTM model for predicting PM2.5 in metro systems
Canyun Yang1, Zhang Kai1, Xinyuan Wang1
1Jiangsu Co-Innovation Center of Efficient Processing and Utilization of Forest Resources, Nanjing Forestry University, Nanjing 210037, China. hongbinliu@njfu.edu.cn.
Environmental Science. Processes & Impacts
|May 12, 2025
Summary
This study introduces a new hybrid model combining Temporal Convolutional Networks (TCN) and Long Short-Term Memory (LSTM) to predict indoor air quality in subways. The model accurately forecasts particulate matter (PM2.5), enhancing passenger health and safety.
Area of Science:
- Environmental Science
- Data Science
- Transportation Engineering
Background:
- Metro systems are crucial for daily commutes, making indoor air quality vital for public health.
- Existing methods for monitoring metro air quality may lack predictive capabilities.
- Particulate matter (PM2.5) is a key indicator of indoor air quality and a significant health concern.
Purpose of the Study:
- To develop a data-driven, soft-measurement model for predicting and optimizing metro air quality metrics.
- To enhance the understanding and management of indoor air environments in public transportation.
- To accurately predict PM2.5 concentrations within subway systems.
Main Methods:
- A novel hybrid model integrating Temporal Convolutional Network (TCN) and Long Short-Term Memory (LSTM) was developed.
- An attention mechanism was incorporated for precise prediction of PM2.5.
- Experiments involved analyzing subway air quality data from Seoul City Hall Station, South Korea.
- TCN model parameters, including residual module size and convolution kernel size, were optimized.
Main Results:
- The proposed TCN-LSTM model demonstrated superior performance compared to baseline models.
- The model achieved a coefficient of determination (R²) of 0.88 on the test set for PM2.5 prediction.
- The hybrid approach effectively captured complex features within the indoor air quality data.
Conclusions:
- The TCN-LSTM model offers a robust and accurate solution for predicting metro indoor air quality, specifically PM2.5.
- This data-driven approach can aid in optimizing air quality management in subway environments.
- The findings contribute to ensuring healthier travel conditions for metro passengers.
