本文把声发射作为序参量描述了岩层运动过程中的自组织性 ,通过 L yapunov指数的计算对岩层运动过程中的混沌性进行了分析 ;提出了确定系统独立变量数目的准则和利用自组织映射神经网络确定独立变量类型的方法 ,最终建立了岩层运动的非线性动力学反演预测方法 ,从而构成了岩层运动非线性动力学预测研究的基本框架 ,这对地震、煤矿冲击地压等地质灾害的预测具有重要参考价值。 In this paper, the acoustic emission is described as a parameter to describe the self-organization in the rock formation movement. The chaos in the rock movement is analyzed by the L yapunov exponent. The criteria for determining the number of independent variables in the rock mass are presented. Neural network to determine the type of independent variables. Finally, a nonlinear dynamic inversion prediction method of rock movement is established, which forms the basic framework of nonlinear dynamic prediction of rock movement. It is of great importance to the geology hazard such as earthquake and coal mine rock burst The prediction has important reference value.