针对综采工作面采煤机智能化截割时存在的煤岩识别精度低等问题，采用红外热像技术，建立了煤岩界面检测实验台。利用煤粉、水泥、沙子、黏合剂等材料分别浇筑全煤、全岩和煤岩混合3种试件，在单光源9 kLux的激励强度下，开展煤岩界面识别研究。实验过程中通过IRBIS3红外分析软件对3种试件检测表面不同时刻的温度场信息进行提取和分析，并建立温度场梯度模型加以研究。研究结果表明:煤、岩两种物质在相同的激励下的温度场特性有明显差异，当存在煤岩混合情况时，检测区域内的温差范围和标准方差会明显增大，可以此作为煤岩混合的一种预测识别依据;对煤岩试件进行主动激励时，检测区域前10 min内温升变化最为明显，温差范围和标准方差增大较快，10~30 min内温度缓慢上升，温差范围和标准方差缓慢增大，30 min后温度相对稳定，温差范围和标准方差无明显变化。激励时间的越长，煤和岩的温度场信息差异就越大，更利于提升煤岩识别的精度。但综合考虑温度场信息差异增长幅度和主动激励能源消耗等问题后，选在主动激励30 min时进行红外图像的提取较为合理;通过提取检测区域内的温度场信息，建立的温度场梯度模型可以较为直观地观察到煤和岩的分布位置及分布趋势，通过对温度梯度模型进行温线划分、煤岩分布划分处理以及建立煤岩分布模拟还原图像后可以准确判定煤岩分布的过渡区域，提升煤岩识别的精度，降低采煤机在截割过程中的截齿磨损程度，有效延长截齿的使用寿命。

Aiming at the problem of low identification accuracy of coal and rock in the intelligent cutting of shearer in a fully mechanized mining face,this paper adopts an infrared thermal imaging technology to establish an experimental platform for active excitation coal rock interface. Three kinds of test pieces of whole coal,whole rock and coal rock were respectively made by using pulverized coal,cement,sand and binder. The coal rock interface identification experi- ment was carried out under the excitation intensity of single light source 9 kLux. During the experiment,the IRBIS3 in- frared analysis software was used to extract and analyze the temperature field information at the different time points of the three test pieces,and the temperature field gradient model was established for the experiment. The results show that the temperature field characteristics of coal and rock under the same excitation are obviously different. When there is coal and rock mixing,the temperature difference range and standard deviation in the detection area will increase signif- icantly,which can be used as a prediction and identification basis for coal rock mixing. When the coal rock specimen is actively excited,the temperature rise in the first 10 minutes at the detection area is most obvious,the temperature difference range and the standard deviation increase rapidly,and the temperature rises slowly within 10-30 minutes, the temperature difference range and the standard deviation are slowly increased. After 30 minutes,the temperature is relatively stable,and the temperature difference range and standard deviation have no significant change. Although the longer the incentive time,the greater the difference in temperature field information between coal and rock,which is more conducive to improving the accuracy of coal rock identification. However,after comprehensively considering the temperature difference of the temperature field information and the active excitation energy consumption,it is reasona- ble to select the infrared image when the active excitation is 30 minutes. By extracting the temperature field information in the detection area,the temperature field gradient model can be compared. The distribution position and distribution trend of coal and rock can be visually observed. By dividing the temperature gradient model into the temperature line, dividing the coal and rock distribution,and establishing the simulated reduction image of coal and rock distribution,the transition area of coal and rock distribution can be accurately determined,and the coal can be upgraded. The accuracy of rock identification reduces the degree of shearer’s pick wear during the cutting process and effectively extends the pick service life.