Coal Spray Rate Prediction Based on Factor Analysis and Neural Network (NN) Algorithm

Abstract

Given the many factors that affect coal bursts as well as complex nonlinear relationships, this study analyzes the main factors that affect coal bursts in coal seams. Central Sulawesi Province. Using NN algorithms to predict coal bursts. This algorithm uses factor analysis to subtract the attributes of the original high-dimensional sample and obtains three common factors to maintain the correlation characteristics of 82.227% of the original data that had 10 sets. factors affecting coal burst rates can be used as training datasets as well as the last five datasets used as testing data and inputted into the rapid miner application to support NN algorithms to predict results. By comparing the prediction results where the NN algorithm has a non-explosion threshold value of 3.194% and an explosion threshold value of -3.230% and is more suitable for predicting coal burst explosions.