Constructing a Gas Explosion Inversion Model in a Straight Roadway Using the GA-BP Neural Network

Area of Science:

• Mine safety engineering
• Computational intelligence
• Explosion dynamics

Background:

• Accurate determination of explosion source parameters is vital for predicting consequences and informing emergency responses.
• Existing methods for explosion source inversion may lack precision, impacting rescue efficiency and safety.
• Roadway gas explosions pose significant risks in underground mining environments.

Purpose of the Study:

• To develop and validate a precise inversion model for determining the location and intensity of roadway gas explosion sources.
• To enhance emergency rescue strategies and accident investigation through reliable data support.
• To improve the overall safety of rescue personnel in post-explosion scenarios.

Main Methods:

• Utilizing a genetic algorithm (GA) to optimize a back propagation (BP) neural network for data inversion.
• Conducting gas explosion experiments to generate simulated overpressure data.
• Implementing the developed inversion model in MATLAB for practical application.

Main Results:

• The GA-BP neural network model demonstrated high precision in locating the explosion source.
• The inversion model achieved high overall accuracy and reliability in predicting explosion disaster parameters.
• The model provides valuable data support for emergency rescue and accident investigation.

Conclusions:

• The developed inversion model offers a significant advancement in accurately assessing roadway gas explosions.
• This methodology provides crucial data for enhancing mine safety and rescue operations.
• The model serves as a reliable tool for both immediate response and post-accident analysis.