Research on safety evaluation of coal mine airflow system based on BP neural network

In coal mine production logistics there are many random factors that result the characters of complexity, non-linear and uncertain of production system. But traditional safety evaluation methods rely on subjective experience, and have a lower evaluating precision. Artificial neural network has better nonlinear mapping ability and high learning ability, which overcomes the deficiencies. Therefore, BP neural network is utilized to establish a safety evaluation model of coal mine airflow system. Firstly, based on the knowledge of coal mine airflow system build an evaluation index system, secondly select reasonable samples of coal mine airflow system as training samples, adjust parameters and add momentum gradient for a better convergence speed, then after a large number of training select the best training network as an evaluation model. Finally, present the application of this model through case analysis, also give reasonable suggestions for coal mine safety production.