Related Experiment Video
Updated: Jul 5, 2025
08:25
BtM, a Low-cost Open-source Datalogger to Estimate the Water Content of Nonvascular Cryptogams
Published on: March 25, 2019
8.1K
A Mine Water Source Prediction Model Based on LIF Technology and BWO-ELM
Pengcheng Yan1,2,3, Guodong Li4, Wenchang Wang2
1State Key Laboratory of Mining Response and Disaster Prevention and Control in Deep Coal Mine, Anhui University of Science and Technology, Huainan, 232001, China.
Journal of Fluorescence
|January 25, 2024
Summary
Laser induced fluorescence (LIF) combined with Beluga Whale Optimization-Extreme Learning Machine (BWO-ELM) accurately identifies coal mine water sources. This advanced method overcomes traditional limitations, enhancing safety and efficiency in mining operations.
Area of Science:
- Geosciences
- Environmental Science
- Data Science
Background:
- Traditional coal mine water source identification methods are slow and risk sample pollution.
- Laser induced fluorescence (LIF) offers high sensitivity and real-time detection for water source identification.
- Extreme Learning Machine (ELM) models face challenges with random weight and bias selection.
Purpose of the Study:
- To develop and evaluate an optimized model for rapid and accurate identification of coal mine water sources.
- To address the limitations of traditional methods and standard ELM by integrating advanced optimization algorithms.
- To enhance the safety and efficiency of coal mine operations by preventing water inrush disasters.
Main Methods:
- Laser induced fluorescence (LIF) technology was used to collect spectral data from mixed water samples.
- Data preprocessing involved polynomial smoothing (SG) and spectral multiple scattering correction (MSC).
- Dimensionality reduction was performed using factor analysis (FA) and linear discriminant analysis (LDA).
- Optimized ELM models, including Beluga Whale Optimization-ELM (BWO-ELM) and Particle Swarm Optimization-ELM (PSO-ELM), were constructed and compared with standard ELM and Long Short Term Memory (LSTM).
Main Results:
- The SG-LDA-BWO-ELM model demonstrated superior performance with a fitting coefficient of 0.99990.
- This optimized model achieved a root mean square error of 0.00041 and a mean absolute error of 0.00021.
- The BWO-ELM model exhibited the best convergence and smallest absolute error among all tested models.
Conclusions:
- The integrated LIF and SG-LDA-BWO-ELM approach provides a highly accurate and efficient method for identifying coal mine water sources.
- This technology is crucial for the timely detection of potential water inrush, significantly improving mine safety.
- The findings support the application of advanced AI-driven spectral analysis for proactive disaster prevention in the mining industry.
