A new study exploring how the technology reduces disaster risk shows that artificial intelligence can predict gas-related events in coal mines in half an hour.
The study of China coal mines compared 10 machine learning algorithms to see which artificial intelligence methods could predict changes in methane gas content 30 minutes in advance and notify users of abnormal conditions. “A Comparative Study of Ten Machine Learning Algorithms for Short-term Prediction of Gas Early Warning Systems” was published in the journal Scientific Reports.
Gas explosions or fires in underground mines pose major risks, and nearly 60% of coal mine accidents are caused by gas gas.
In 2020, coal production accounted for 46% of the world’s total, and more than 3200 coal mines with high gas content were at outstanding risk levels.
Niusha Shafiabady, author and adjunct associate professor at the School of Science and Technology at Charles Darwin University (CDU), said the results showed that four of the 10 machine learning algorithms produced the best results.
Associate Professor Shafiabady said: “Linear regression is one of the most effective algorithms and has better performance than other algorithms in terms of short-term prediction.”
“Random forests often exhibit statistically low error performance and achieve the highest prediction accuracy. Support vector machines perform well and have shorter computing times on small datasets, but as the dataset size increases, it requires too much training time.
“The results of this study will help the coal mining industry reduce the risk of accidents such as gas explosions, ensure worker safety, and improve its ability to prevent and mitigate disasters that may not only cause casualties, but also cause economic losses.”
The research was conducted in collaboration with Charles Darwin University, University of Technology Sydney, Catholic University of Australia, Shanxi Normal University and Central Queensland University.
Associate Professor Niusha Shafiabady, a researcher at the Peter Farber School of Business at the Catholic University of Australia, said these results have multiple applications.
“This method applies to all coal mines, and the same principles apply to aerospace, oil and gas, agriculture and other industries,” she said.
“This is an example of an application where artificial intelligence can be used to save lives and mitigate health and safety risks.”
Previous research by Associate Professor Shafiabadi has found that more monitoring of wind, gas density and temperature in coal mines can also help reduce disaster risks.
Original text:https://techxplore.com/news/2024-09-ai-disasters.html
More information: Robert MX Wu et al., Comparative study of ten machine learning algorithms for short-term prediction of gas warning systems, Scientific Report (2024). DOI：10.1038/s41598-024-67283-4
Journal information: Scientific reports
Provided by Charles Darwin University
Oil tubing: