论文部分内容阅读
为了有效地对隧道施工进行安全状态评估,建立了以人-机-环-管理系统为基础的隧道施工安全状态评估指标体系,构造了基于粗糙集-模糊评判-神经网络的隧道施工安全状态评估模型。该模型通过粗糙集约简输入变量,提炼学习样本,再利用神经网络对其进行训练和评价,提出了使用层次分析法和模糊数学的方法对隧道施工安全整体综合评判,得到的评价值作为神经网络训练目标值的方法。实际结果表明,通过使用该模型方法,神经网络训练的条件属性由原来的17个变成6个,训练周期由原来的2992次减少为1637次,泛化能力、安全状态评估的结果都优于约减前,能够对隧道施工安全状态做出有效的评估结论。
In order to assess the safety status of tunnel construction effectively, the evaluation index system of tunnel construction safety based on MAN-machine-ring-management system is established, and the safety assessment of tunnel construction based on rough set-fuzzy judgment-neural network model. The model simplifies the input variables by rough set, extracts the learning samples, and then uses neural network to train and evaluate them. Based on AHP and fuzzy mathematics, this model puts forward the overall evaluation of tunnel construction safety. The obtained evaluation value is used as neural network Method of training target value. The actual results show that by using this model method, the condition attributes of neural network training are changed from 17 to 6, and the training period is reduced from 2992 to 1637. The results of generalization ability and safety status assessment are better than About to reduce before the tunnel construction to be able to make an effective assessment of the state of safety conclusions.