煤炭在我国能源结构中占据着核心地位，但煤矿开采安全深受煤自燃灾害的影响，其引发的煤矿火灾不仅造成环境污染、危害从业人员的生命财产安全，还会加剧其它灾害的影响。因此，煤自燃的探测和预警对矿井工作安全、大气环境保护和国家能源持续发展具有十分重要的研究意义和现实价值。煤自燃的探测原理通常是采用设备对可能揭示煤火系统物质和结构的物理化学参数进行检测，再通过数据反演推测煤自燃的分布范围及发展程度。根据系统中水、热、力、化、电等的变化特征，出现了基于不同特征指标的探测方法，包括测温法、测气法、测氡法、遥感法和电磁法等。与其他方法相比，电磁法因为具有探测范围广、可靠性高、效果好等明显的优势而受到了广泛关注。

在煤自燃探测过程中，不同温度下煤岩分层结构中电磁波的传播规律是电磁法的核心问题。因此，本文首先对升温过程中煤介质电学特性的变化规律进行研究，得到温度-频率-电磁参数的作用模型；其次，对典型煤岩分层结构的时域反射特性进行分析，建立温度与电磁场的作用关系；最后，根据煤岩分层结构的瞬态电磁响应，对自燃阶段进行划分。基于此，本文开展的主要工作如下：

（1）研究了不同煤种的温度—频率—电磁参数作用模型。提出电磁参数随温度变化的频率修正因子，建立温度、频率和电磁参数的三维模型，研究宽频带内不同温度下电磁参数变化规律，为煤岩分层结构瞬态电磁响应特性的研究提供必要基础；

（2）结合广义传播矩阵法和傅里叶变换，可以一次计算宽频带下复杂煤岩分层结构的反射回波，而无需考虑数值建模带来的计算困难。基于此，分析了煤层结构参数对反射系数特性的影响，以及自燃过程中煤层温度对时域反射回波特性的影响；

（3）基于斯皮尔曼相关系数法对煤岩分层结构的瞬态电磁响应特征与煤燃烧温度之间进行相关分析，研究了反射波峰值变化率作为燃烧阶段识别指标的实用性。基于此，对煤自燃的燃烧阶段进行了划分。

Coal occupies a core position in China's energy structure, but in recent years, coal mining safety has been deeply affected by spontaneous coal combustion fires. Coal fires triggered by spontaneous coal combustion not only directly jeopardize the life and property safety of coal miners, but also exacerbate their other disasters and pollute the environment, so early warning of spontaneous coal combustion disasters has always been a hot issue in coal mine safety. According to the change characteristics of water, heat, force, chemical, and electricity in the coal fire system, the detection methods based on different characteristic indexes have appeared, including temperature measurement method, gas measurement method, radon measurement method, remote sensing method, and electromagnetic method. Compared with other methods, the electromagnetic method with obvious advantages such as wide detection range, high reliability, and good effect has received wide attention.

The continuous high-temperature burning of coal underground constitutes a complex and dynamically changing coal fire system, and the temperature and electromagnetic wave frequency, as the two main factors, make the propagation of electromagnetic waves in the layered structure of coal and rock more complicated. Therefore, in this paper, firstly, the changing law of the electrical characteristics of the coal medium in the warming process is investigated, and the role model of temperature-frequency-electromagnetic parameters is obtained; secondly, the characteristics of the time-domain reflected waves of typical coal-rock layered structures are analyzed, and the interaction relationship between temperature and electromagnetic field is obtained; finally, the stage of spontaneous combustion is classified according to the transient electromagnetic response of the coal-rock layered structures. Based on this, the main work carried out in this paper is as follows:

(1) The temperature-frequency-electromagnetic parameter interaction model of different coal types is studied. The frequency correction factor of electromagnetic parameters with temperature is proposed, and a three-dimensional model of temperature, frequency, and electromagnetic parameters is established to study the variation rule of electromagnetic parameters at different temperatures in a wide frequency band, to provide the necessary basis for the study of transient electromagnetic response characteristics of coal-rock layered structure;

(2) Combining the generalized propagation matrix method and the Fourier transform, the reflection echoes of the complex coal-rock layered structure in a wide frequency band can be calculated simultaneously without considering the computational difficulties caused by numerical modeling. Based on this, the effects of coal seam structural parameters on the reflection coefficient characteristics; and the effects of coal seam temperature on the time-domain reflection echo characteristics during spontaneous combustion are analyzed;

(3) Based on Spearman's correlation coefficient method, a correlation analysis is carried out between the transient electromagnetic response characteristics of the coal-rock layered structure and the coal combustion temperature, and the utility of the peak rate of change of the reflected wave as an indicator for the identification of combustion stages is investigated. Based on this, the combustion stages of spontaneous coal combustion are divided.