Neural Network Model Based on Genetic Algorithm for Predicating Mining Subsidence in Multi-fault Areas

摘要：

In order to make neural network model better used in multi-fault areas subsidence prediction, BP neural network was first improved, which was then combined with the genetic algorithm to search the optimal structure of BP neural network, accelerate network convergence, avoid local minimum, and establish GA-BP model for surface subsidence prediction. Combined with the measured data samples of the surface subsidence in several mining areas of ,Nanye coal mine, training and learning on this neural network model were carried out, and then the surface subsidence in the fault areas was predicted with this model. The results show that the neural network model has such advantages as fast convergence and higher forecast accuracy. It proves to be a practical approach in predicting mining subsidence in multifault areas.

关键词： Neural network BP algorithm Genetic algorithm Mining subsidence

作者: LI Tingchun XING Xueyang SHI Zhaoxia

作者单位: Shandong Provincial Key Laboratory of Civil Engineering Disaster Prevention and Mitigation, Shandong University of Science and Technology, Qingdao, Shandong,China,266510

会议类型: 国际会议

会议名称: 2010 International Conference on Mine Hazards Prevention and Control(第二届矿山灾害预防与控制国际学术会议 ICMHPC)

会议地点: 青岛

会议语种:英文

页码: 474-480

在线出版日期: 2010-10-15（万方平台首次上网日期，不代表论文的发表时间）