本文通过阅读大量文献资料，在了解岩浆岩侵入体三维地震解释发展现状的基础上，归纳分析了岩浆岩侵入体的地质特征和物性特征， 阐述了岩浆岩侵入体的常规地震解释和地震属性解释，针对常规解释与属性解释存在的问题，采用灰色关联分析方法解决了属性参数优化问题， 在此基础上依据具体的研究区实例，设计出岩浆岩沿各煤层顶底板侵入的理论正演模型，论述了用人工神经网络方法实现岩浆岩侵入体预测的可行性。 最后经用某研究区8煤层与10煤层三维地震属性数据进行验证，结果比较理想。 该论文是在岩浆岩沿煤层顶底板顺层侵入矿区，将地震属性技术、地震属性优选方法和BP神经网络预测技术结合起来， 解决煤矿岩浆岩侵入范围圈定的应用技术研究，其成果可为矿井开拓布置提供准确的地质资料，给煤矿生产带来巨大的经济效益。

This text is described the common seismic explanation and seismic attribute explanation of magmatic intrusive mass by reading a large amount of documents and materials and analysing geological characteristics and thing characteristics of magmatic intrusive mass, on the basis of understanding the current development situation of magmatic intrusive mass’s 3D seismic explanation. The text is adopted analytical methods of the grey related to solve the optimization problem of attribute parameters for the problems of existing common explanation and attribute explanation.According to instances in reconnaissance area, Text has been expounded the fact that the feasibility of predicting Magmatic intrusive mass with the method of the artificial neural network through designing perform model of Magmatic intrusive mass along one layer of forms invasion of every coal seam. Finally, it is more ideal result that was verified by 3D seismic attribute data about 8 coal seam and 10 coal seam in one colliery. Thesis has been combined the seismic attribute and optimum seeking methods with BP neural network technology to solve range of magmatic intrusive mass in mining area, its achievement can offer the accurate geological materials for opening and assigning in the mine, making the enormous economic benefits to the production of the colliery.