Research On Mine Water Inrush Prediction Based On Extreme Learning Machine And Route Optimization

Coal plays an important role in national energy development strategy as the most important source. In China, complex geological structure and coal mine water inrush lead to serious economic losses for normal production, and even endanger the safety of underground personnel. In the process of coal mining, predicting coal mine water inrush and designing personnel escape routes are effective means to ensure mine safety.Mine water inrush is the result of the many factors. These factors are interrelated which make the relationship between water inrush and influencing factors become very complex nonlinear relation. So we can’t build a fast and accurate model for prediction of water inrush by using traditional mathematical models. Therefore, this paper propose a prediction model by studying on the existing water inrush methods, which is based on PSO-ELM, and PSO-WELM. In this paper, we prove the PSO-WELM is more in line with the forecast requirements of coal mine water inrush prediction based on the verification and analysis of the two models using the classical UCI datasets, and then apply the model in coal mine water inrush prediction.In application of the PSO-WELM coal mine water inrush prediction model. Firstly, we select the main factors affecting coal mine water inrush based on the mechanism of coal mine water inrush. Then collecting a large number of historical data about coal mine water inrush and breaking down the data into training set and test set. Then training and testing the coal mine water inrush prediction model by using these data to build the coal mine water inrush prediction model to forecast the possibility and the type of water inrush.In addition, this paper also do some research about the escape routes once the coal mine water inrush happened, weither it is caused by natural factors or by using water inrush prediction model. No matter what accident happened, we need to organize workers evacuate to a safe location quickly. The underground environment is complex and harsh, once the water inrush occurred the underground environment will become more complex. Therefore, how to make the underground personnel escape along the best of route to safety position with the fastest speed is another focus in this paper.Choosing a reasonable escape route, we need to consider the impact of water and other factors on the roadway. In this case, the distance between two points is no longer the absolute distance, but the relative distance, this paper introduces the concept of equivalent length of the roadway to indicate the extent of the prevailing difficult. In this paper, we proposed the multi-optimal paths of coal mine floods based on account of the D-K algorithm based on the Dijkstra algorithm and the K-shortest path algorithm, which can obtain multiple best escape route from the site of the incident to the coal mines all escape wellheads. It can not only ease congestion personnel, but also evacuat personnel timely and orderly to a safe location. In this paper, combined with the actual example of mine, we prove the advantages of D-K algorithm used in coal mine flood escape route of choice by comparing D-K algorithm with Dijkstra algorithm and the Kshortest path algorithm.