利用激光诱导等离子体光谱(LIPS)技术实现对燃煤电厂输煤管中煤质的在线检测对于提高锅炉燃烧效率具有重要的现实意义，但对煤中氧含量的测量却是个难题。提出了一种新的数据处理方法，主要包括内标法、最佳分析线选择法、温度校正法及多线法等，来提高对煤中氧含量定量分析的准确性和重复性。实验中先通过8组煤样获得了煤中氧的定标方程，然后又利用其它6组煤样来验证该定标方程的准确性。实验结果表明，利用本数据处理方法对煤中氧含量测量的绝对误差小于1.1%，多次测量的相对标准偏差(RSD)小于5.9%，显示了较高的测量精度和重复性。

To improve the combustion efficiency, it is important for coal-fired power plants to achieve on-line coal quality analysis in coal pipe by using laser-induced plasma spectroscopy (LIPS). However, how to determine the total content of oxygen in coal becomes a major obstacle. A new data processing method including the methods of internal normalization, the optimal emission-line selection, temperature correction and multi-line analysis has been proposed to increase the measurement accuracy and repeatability. In the experiment, the calibration equation of total oxygen is obtained with eight coal samples firstly, and then the accuracy of the equation is verified with another six coal samples. Experimental results show that the absolute measurement error for the total oxygen with this data processing is within 1.1% and the relative standard deviation of multiple measurements is within 5.9%, which show the relative high accuracy and repeatability.