Establishment and Application of Thin Coal Seam Mining Method Prediction Model Based on Improved Neural Network

Thin coal seam mining is one of the key problems in coal mining,being able to predict thin coal seam mining method is an essential aspect in coalmine sustainable development.However,the traditional methods concerning mining method selection is hindered by multi-objective and non-continuous problem.The thin coal seam mining method prediction model based on improved neural network (TCSMMPM-INN) was established through analysis of thin coal seam mining methods using an improved neural network algorithm designed to train the network and using an preferred inspection method to resolve the over-fitting problem,the network test and field application results show that: the predicted result of thin coal seam mining method is 100%,the error values of working face yield are from 2.2% to 6.5%,the error values of ergonomics are from 2.3% to 8.7%,this model can provide a new direction for research into predicting the likelihood of thin coal seam mining method selection within a coalmine and will have a broad application prospect in coal mining.