299 / 2018-05-02 12:21:24

Prediction of quarterly rock burst accident in Coal Mine Based on combination model

rock burst,accident,prediction method,combination model

矿山安全科学理论与应急救援 > 矿山安全信息化

Based on the information of rockburst accidents in coal mine in China from 2006 to 2015, taking the number of quarterly accidents data in Central China and Southern China region as the research object, this paper studies the rockburst accidents in coal mine in the quarter of 10 years. The prediction model uses the pretreatment method of HP filtering and data normalization, and the number of accidents of central China and Southern China region is predicted using combined ARMA- BP neural network model. compared to ARMA, ARIMA, BP neural network prediction model, the results show that the prediction accuracy of the ARMA-BP combination model based on HP filter is highest, which is suitable for the prediction of the data.