A novel prediction model for the degree of rescue safety in mine thermal dynamic disasters based on fuzzy analytical hierarchy process and extreme learning machine

Abstract

Considering the rapid development, uncertain situations and prediction difficulty of mine thermal dynamic disasters (MTDDs), this paper combines the fuzzy analytical hierarchy process (FAHP) and extreme learning machine (ELM) into a prediction model to quantify the degree of MTDD rescue safety in a fast and accurate manner. Firstly, a static FAHP model was constructed by the Delphi, AHP, and fuzzy comprehensive evaluation (FCE) to assess various MTDD rescue cases, quantify the exact degree of rescue safety and provide sample data for real-time prediction based on artificial intelligence. Then, the improved ELM algorithm was introduced to design a dynamic prediction model for the safety of various MTDD rescue cases. Through case study, it is confirmed that the prior knowledge of experts was fully utilized by the model, outputting accurate, rational and reliable sample data. Moreover, the ELM algorithm enabled the intelligent extraction and rapid inference of sample data features. The prediction results are accurate, stable, and highly reliable, revealing that our model is a desirable tool for the real-time prediction of safety in mine rescue operations. The research findings provide new insights into the prediction of mine rescue safety.