Multi-Level Forecasting Model of Coal Mine Water Inrush based on Self-Adaptive Evolutionary Extreme Learning Machine

This paper constructs a forecasting model for coal mining water inrush from the floor through the analysis and careful study of the mechanism of water inrush in coal mining and the Self-Adaptive Evolutionary Extreme Learning Machine (SaE-ELM) which obtains self-learning, generalization performance and speediness is used. In SaE-ELM, the network hidden node parameters are optimized by the self-adaptive differential evolution algorithm, whose trial vector generation strategies and their associated control parameters are self-adapted in a strategy pool by learning from their previous experiences in generating promising solutions. Large amounts of historical data of mining water inrush is collected and the main controlling factors are extracted as sample data to train and test the forecasting model by the SaE-ELM, which can forecast both the existence of a water inrush from the floor and the level of the water inrush. Experiments are given to prove that the proposed method reduces the time of model construction and computation, and improves the speed and accuracy of the forecast of coal mining water inrush.