摘要
This study was conducted to establish a Support Vector Machines(SVM)-Markov Chain prediction model for prediction of mining water inflow. According to the raw data sequence, the Support Vector Machines(SVM) model was built, and then revised by means of a Markov state change probability matrix. Through dividing the state and analyzing absolute errors and relative errors and other indexes of the measured value and the fitted value of SVM, the prediction results were improved. Finally,the model was used to calculate relative errors. Through predicting and analyzing mining water inflow, the prediction results of the model were satisfactory. The results of this study enlarge the application scope of the Support Vector Machines(SVM) prediction model and provide a new method for scientific forecasting water inflow in coal mining.
This study was conducted to establish a Support Vector Machines （SVM）-Markov Chain prediction model for prediction of mining water inflow. According to the raw data sequence, the Support Vector Machines （SVM） model was built, and then revised by means of a Markov state change probability matrix. Through dividing the state and analyzing absolute errors and relative errors and other indexes of the measured value and the fitted value of SVM, the prediction results were improved. Finally, the model was used to calculate relative errors. Through predicting and analyzing mining water inflow, the prediction results of the model were satisfactory. The results of this study enlarge the application scope of the Support Vector Machines （SVM） prediction model and provide a new method for scientific forecasting water inflow in coal mining.
