Related Experiment Videos
Improving Road Traffic Forecasting Using Air Pollution and Atmospheric Data: Experiments Based on LSTM Recurrent
Faraz Malik Awan1, Roberto Minerva1, Noel Crespi1
1Telecom SudParis, Institut Polytechnique de Paris, CNRS UMR5157, 91000 Evry, France.
Sensors (Basel, Switzerland)
|July 9, 2020
Summary
This study enhances traffic flow forecasting by incorporating air pollution and atmospheric data. Integrating these factors with traditional traffic data improves prediction accuracy for smart city management.
Area of Science:
- Environmental Science
- Transportation Engineering
- Data Science
Background:
- Accurate traffic flow forecasting is crucial for smart city management and driver navigation.
- Existing research often predicts air pollution using traffic data, but not vice versa.
Purpose of the Study:
- To develop an improved traffic forecasting approach by integrating air pollution and atmospheric parameters.
- To investigate the relationship between traffic intensity, air pollution, and atmospheric conditions.
Main Methods:
- Utilized air pollution data (CO, NO, NO2, NOx, O3) and atmospheric parameters (pressure, temperature, wind direction/speed).
- Collected data from Madrid, Spain's open data portal.
- Employed a Long Short-Term Memory (LSTM) recurrent neural network (RNN) for forecasting.
Main Results:
- Preliminary experiments confirmed a relationship between traffic intensity, air pollution, and atmospheric parameters.
- The addition of these environmental factors is hypothesized to enhance traffic forecasting accuracy.
Conclusions:
- Integrating air pollution and atmospheric data offers a novel approach to improve traffic flow predictions.
- This method holds potential for more effective smart city traffic management.