中文摘要：

利用实验测定的电磁辐射信号时间序列，用双向差分原理反导出一个非线性常微分方程；以其为微分动力核，运用动力系统数据机理自记忆模式构造自记忆方程并求出自记忆系数；利用该方程预测未来电磁辐射信号的变化，并与现场测定对比分析，用误差分析和距平分析法验证该模型正确性和预测准确率。实例表明：该自记忆模型预测与实测结果是一致的，相对误差均在6．7852％左右，距平符合率为90％；自记忆方法能有效应用于煤与瓦斯突出电磁辐射动态预测中；该模型与电磁辐射预测方法的有机结合能有效地提高预测准确性，从而为煤与瓦斯突出电磁辐射预测技术提供了一种新的研究途径。

英文摘要：

On the basis of experimental research on prediction of coal or gas outburst by means of electromagnetic radiation （EMR） method which is a kind of non-contacting forecasting methods, according to self-memory theorem in dynamic process, the self-memorization model of coal and gas outburst＇ prediction model by means of electromagnetic emission （EME） method is researched. Firstly, by use of an observed time data series of in-situ EME signals, a nonlinear ordinary differential equation based on the bilateral difference principle is retrieved. Taking the above nonlinear ordinary differential equation as a dynamic kernel, with the self-memorization principle a forecast model can be established, which is called the Databased Mechanistic Self-memory Model （DAMSM）, and the self-memory efficiency is also given. Finally, the above forecasting equation is used to predict the EME data in the future, and the prediction fitting value is compared with the practical data. Some computing cases are given which show that the forecasting aceuraey of self-memorization model is satisfactory, and the organic combination of EME method and self- memorization model can predict the coal and gas outburst efficiently. In this work, a new research idea and method is provided for prediction of coal and gas outburst based on EME prediction method.