基于GIS的岩爆倾向性模糊自组织神经网络分析模型

APPLICATION OF GIS-BASED NEURAL NETWORK WITH FUZZY SELF-ORGANIZATION TO ASSESSMENT OF ROCKBURST TENDENCY

岩爆倾向性是评价地层深部工程安全性的一个十分重要的指标之一。研究方向把常规的研究岩爆倾向性的方法与GIS技术结合起来,利用GIS的空间数据分析技术和模糊自组织神经网络,对评价岩爆倾向性的多源信息进行加工处理,建立了一个基于GIS技术的岩爆倾向性模糊自组织神经网络模型。

Rockburst tendency is one of the important indices for the safety evaluation of deep mining engineering. The accuracy of rockburst tendency analysis and assessment is utterly dependent on the reliability of raw mechanics data and the rationality of mathematical models. To upgrade the reliability of data, geographical information system (GIS) is used to support management of multi-resource raw data, perform data processing, refine map layer for the factors needed in analysis, and easily divide the region into units for analysis. Then, via secondary programming based on GIS, the neutral network model with fuzzy self-organization is combined with GIS and conventional methods tightly for rockburst tendency assessment, i.e., GIS takes the role of providing input data for the neutral network with fuzzy self-organization as well as processing its analysis results and outputting them in forms of maps. The feasibility and practical methods to integrate neutral network with fuzzy self-organization and GIS are discussed, in the light of a deep mine in China.