Assessment And Application Research Of BP Neural Network On Coal Mine Safety Culture Maturity

Safety culture is the unit of shared safety values,attitudes,moral and behavioral norms of group.Studies show that safety culture can effectively affect person's safety behavior and self-discipline from the spiritual field,also more persistence compared to other safety management means.Statistical data indicates that almost 97% coal mine accidents are caused by human factors or unsafe behavior related human factors.Thus,how to improve personnel quality related to safety in coal mines,enhance the level of safety culture as a whole and improve the coal mine's production,is of vital importance.It can also help with accident prevention and cannot be underestimated.In order to evaluate coal mine safety culture effectively,introduce the concept of safety culture maturity,the development of safety culture maturity is divided into five stages.A two-dimension safety culture model is built using BP neural network which is applicable to coal mine's safety culture maturity.Specific work is as follows.(1)On the basis of the definition of maturity,set up the five-step maturity model for coal mine safety culture,include original,reusable,defined,management,optimizing continuously,then refine the characteristics.Using expert investigation method to coillect the points,thus the first dimension of the model is formed.(2)In order to ensure the accuracy and reliability of the evaluation indicators,considering four aspects of security factors in system risk,as people,machine,environment,and management according to the characteristics of the coal mine safety.Finally establish an 18-factors evaluation system to evaluate the afety culture maturity.With the relavant experts' score,build the second level for the model.(3)With BP neural network,under the platform of Matlab R2014 a,coal mine safety culture maturity evaluation model is to build.This paper selects a certain number of training samples and use random interpolation method to expand the samples.The neural network can maintain assessment knowledge and experience after training.Select three companies as the text data for the improvement of convergence speed and stability of the BP neural network.The test result and example of coal mine indicates that compared with other evaluation methods,this model used to illustrate the method is effective.