Safety Warning Model of Coal Face Based on FCM Fuzzy Clustering and GA-BP Neural Network

Risk and security are two symmetric descriptions of the uncertainty same system. If risk early warning is carried out in time, capability system can be improved. A safety model based on fuzzy c-means clustering (FCM) back-propagation neural network was established, a genetic algorithm introduced to optimize connection weight other properties network, so as construct coal mining face. The applied face Shandong, China, with 46 groups data samples. Firstly, original were clustered by FCM, input space divided, samples into three categories. Then, used for training prediction. optimization trained verified many times. results show that realize prediction condition working face, performance optimized better than traditional artificial model, higher accuracy convergence speed. established method provide reference basis prediction, management mine production safety, discover hidden danger accident soon possible, eliminate time reduce probability maximum extent.