Risk Assessment of Deep Coal and Gas Outbursts Based on IQPSO-SVM.

Area of Science:

• Mining Engineering
• Artificial Intelligence
• Computational Science

Background:

• Coal and gas outbursts pose significant risks to deep mine safety.
• Accurate risk assessment is crucial for preventing mining accidents.

Purpose of the Study:

• To develop an advanced risk evaluation method for deep coal and gas outbursts.
• To address the challenges of high-dimensionality, nonlinearity, and small sample sizes in risk assessment.

Main Methods:

• A novel Improved Quantum Particle Swarm Optimization Support Vector Machine (IQPSO-SVM) model was developed.
• IQPSO was employed to optimize Support Vector Machine (SVM) parameters, overcoming limitations of standard PSO and QPSO.
• The algorithm enhances global and local search capabilities for improved data fitting.

Main Results:

• IQPSO-SVM demonstrated significantly higher accuracy in risk assessment compared to standard SVM, PSO-SVM, and QPSO-SVM.
• The proposed method effectively improves parallelism, stability, robustness, and model generalization.

Conclusions:

• IQPSO-SVM offers a new approach for predicting and preventing deep coal and gas outbursts.
• This study provides a valuable reference for evaluating complex, high-dimensional, and nonlinear problems in various scientific fields.
• The findings support enhanced safety management in deep coal mines.