基于RS-ANN的煤矿安全控制

针对目前煤矿安全管理的现状，提出利用粗集-神经网络对煤矿安全进行控制.模型在基于人-机-环境理论基础上，全面分析了影响煤矿安全的因素，利用基于蚁群算法的粗糙集属性约简对安全因素进行分析.将粗糙集方法融入神经网络实现优势融合可以去掉冗余输入信息、减小神经网络构成系统的复杂性. 提高容错及抗干扰的能力.在此基础上，利用人工神经网络的预测功能，预测影响煤矿安全的关键因素，并根据预测结果提出有针对性的安全技术措施加以防范.用同一组数据比较该方法与典型BP网络的预测效果，结果表明该方法明显优于BP网络.

Coal mine safe control model based on RS-ANN

According to the current management of coal mine safety status，a coal mine safe control model based on rough sets-artificial neural network（RS-ANN） was established. Based on Man-Machine-Environment theory, safe factors that effect the realization of coal mine safe aim were obtained. Combining theRough sets theory that based on the ant colony algorithm with the Neural Networks, the super combination can realize to delete the superfluous inputting information, reduce the complexity and improve the interfere resistance. Therefore, a basic thought and specific method to set up Rough sets-Neural Network system to control the coal mine safe is presented, which introduce rough sets reduction method and obtain the mini safe factor in thehistoric data.At last, the neural network system can control the expect aim of coal mine safe management. The forecast results show that this approach is better than the typical BP NN with the same data．