高承压底板突水灾变演化过程的多源异构信息识别

This study research on the multi-source heterogeneous information recognition, fusion and early warning method for the catastrophe evolution of floor water inrush caused by coal mining above confined water with high pressure. Physical simulation and numerical simulation will be used in the study. More specifically, combined with field test data, a number of measurable precursor information of typical water inrush disasters and water inrush simulation test will be obtained. And then, inverse analysis method will be used to select and inverse the information, the incremental relation criterion between the catastrophe evolution of water inrush channel and multi-information will be determined. According to the D-S evidence theory and BP neural network technology, a compositive identification, information fusion, criterion and warning method will be proposed. Our experiment and systematic research will lay a theoretical foundation for the short-term prediction and effective control of coal mining above confined water with high pressure.

本申请课题以高承压水上采煤的特殊开采条件为例，重点研究底板突水的灾变演化过程中多源异构前兆信息的识别、融合、预警方法。本课题将采用物理模拟、数值模拟等手段，结合现场实测数据，基于多传感器获取典型突水事故及突水模拟试验中的若干可测前兆信息；利用反分析法，通过参数筛选、优化及反演，建立基于参数优选和矫正方法的突水灾变演化与多源信息的渐进式关联准则；基于D-S证据理论和BP神经网络技术，提出多源异构突水前兆信息的综合识别模型、融合方案、判别准则及预警方法。研究成果可为高承压底板突水的临突监测预报及有效控制提供理论依据。

本课题在开展典型矿区实地调研的基础上，总结提出了具有代表性的三种底板高承压开采条件下的水文地质结构模式；在此基础上，开展了底板高承压开采条件下底板采动裂隙形成及演化过程的室内数值模拟、现场实测，建立了若干可测、有效信息源的监测方案，通过对前兆信息的监测与有效性分析，开展了室内高承压底板突水模拟试验三台，进一步对采动通道形成、演化过程直至突水事故发生的整个灾变过程的包括底板应力场、水动力场等方面的若干有效前兆信息进行了监测，获取了若干具有代表性的有效突水前兆监测信息源。在此基础上，结合大型突水案例资料，确定了有关矿井突水渗流场、力学场、地球物理场以及水化学场的突水前兆监测信息的综合等级，提出了矿井突水的单因素、多因素综合监测预警的识别模型以及矿井突水水源的快速识别、定量构成判别模型。