Abstract:
The data of existing gas emission prediction methods of stop working face are mostly based on gas concentration sequence of single sensor in stope working face, and these methods can not record position of monitoring point in process of continuous advancement of the working face.In view of above problems, a method that used BP neural network model to predict gas emission in the working face was proposed, which was based on data of gas concentration sequence data of monitoring point of sensor and actual advance distance on stope working face. The method uses gas source identification method of the working face to analyze variation law of gas emission of in goaf and coal wall respectively; and uses BP neural network prediction method to predict average daily gas emission combining with characteristic values of variation law of gas emission of in goaf and coal wall. The example application verifies correctness of the method.