Particulate matter concentration from open-cut coal mines: A hybrid machine learning estimation
Chongchong Qi1, Wei Zhou2, Xiang Lu2
1School of Civil, Environmental and Mining Engineering, University of Western Australia, Perth, 6009, Australia; School of Resources and Safety Engineering, Central South University, Changsha, 410083, China.
Environmental Pollution (Barking, Essex : 1987)
|April 14, 2020
Summary
A hybrid machine learning model accurately predicts particulate matter (PM) concentrations from coal mines. This approach aids in assessing air quality risks and designing effective dust control strategies.
Area of Science:
- Environmental Science
- Data Science
- Mining Engineering
Background:
- Particulate matter (PM) emissions pose significant environmental challenges in the coal mining industry.
- Accurate PM concentration prediction is crucial for implementing effective control strategies.
Purpose of the Study:
- To develop and validate a hybrid machine learning model for precise PM concentration estimation.
- To optimize the Random Forest (RF) model using Particle Swarm Optimization (PSO) for hyper-parameter tuning.
Main Methods:
- A hybrid RF-PSO model was developed to predict PM2.5, PM10, and Total Suspended Particulate (TSP) concentrations.
- Meteorological data (wind direction, speed, temperature, humidity), noise level, and prior PM concentration were used as inputs.
- The model was applied to data from the Haerwusu Coal Mine in northern China.
Main Results:
- The RF-PSO model demonstrated high accuracy in estimating PM concentrations.
- Pearson correlation coefficients between estimated and measured data were 0.91 (PM2.5), 0.84 (PM10), and 0.86 (TSP).
- PM concentration at 5 minutes prior was the most significant factor, followed by humidity, temperature, noise level, wind speed, and wind direction.
Conclusions:
- The study presents an efficient and accurate method for PM concentration estimation in open-cut coal mines.
- This predictive capability is essential for assessing atmospheric quality risks and informing dust control technique design.
