激光诱导击穿光谱精确测定燃煤工业分析指标的研究

发布时间：2018-01-23 10:36

本文关键词： 激光诱导击穿光谱燃煤工业分析指标光谱拟合支持向量机　出处：《光谱学与光谱分析》2017年10期 　论文类型：期刊论文

【摘要】：燃煤工业指标的在线精确分析对于指导燃煤工业优化生产、降低燃煤煤耗至关重要。利用激光诱导击穿光谱(LIBS)分析燃煤煤质时,因受我国复杂多样煤种所导致的"基体效应",测量精度有待提高。实验中对激光诱导燃煤等离子体光谱至燃煤工业分析指标转化过程中的光谱预处理和定标建模方法进行了优化选择。实验结果表明,利用单/多峰Lorentzian光谱拟合计算谱线强度相比于传统计算方法,谱线强度RSD均值可由12.1%降至9.7%;对于核函数参数寻优,相比于网格参数(Grid)和遗传算法(GA),粒子群算法(PSO)的平均绝对误差(MAE)最小;采用PSO参数寻优式支持向量机(SVM)回归建模的预测均方根误差(RMSEP)小于偏最小二乘回归分析法(PLS);采用单/多峰Lorentzian光谱拟合方法和PSO参数寻优式SVM回归建模,对燃煤工业分析指标预测的平均绝对误差(AAE)为:灰分为16%~30%时AAE为1.37%,灰分大于30%时AAE为1.77%,发热量为9~24 MJ·kg-1时AAE为0.65 MJ·kg-1,挥发分低于20%时AAE为1.09%,挥发分大于20%时AAE为1.02%。
[Abstract]:On-line accurate analysis of coal industry index is very important to guide coal industry to optimize production and reduce coal consumption. Laser induced breakdown spectroscopy (LIBS) is used to analyze coal quality. "Matrix effect" caused by complex and diverse coal types in China. The measurement accuracy needs to be improved. In the experiment, the spectral pretreatment and calibration modeling method in the process of laser induced coal-fired plasma spectrum conversion to coal-fired industrial analysis index are optimized and selected. The experimental results show that. Compared with the traditional method, the line intensity calculated by single / multi-peak Lorentzian spectrum fitting can be reduced from 12.1% to 9.7. For parameter optimization of kernel function, the mean absolute error (mae) of particle swarm optimization (PSO) is the smallest compared with grid parameter (Grid) and genetic algorithm (GA). The root-mean-square error (RMS) based on PSO parameter optimization support vector machine (SVM) regression model is less than that by partial least square regression analysis. Single / multi-peak Lorentzian spectral fitting method and PSO parameter optimization SVM regression model were used. The average absolute error of the prediction of the coal industry analysis index is as follows: the AAE of ash is 16 ~ 30 and the AAE is 1.77% when ash is greater than 30. The AAE was 0.65 MJ 路kg-1 when the calorific value was 9 掳24 MJ 路kg-1, and the volatile content was lower than that at 20 MJ 路kg-1, which was 1.09% MJ 路kg-1. AAE is 1.02 when volatile matter is greater than 20.
【作者单位】：山西大学激光光谱研究所量子光学与光量子器件国家重点实验室;极端光学协同创新中心山西大学;
【基金】：国家(973计划)项目(2012CB921603) 国家自然科学基金项目(61475093,61378047,61575113,61404104,61275213) 山西省科技重大专项(MD2016-01) 山西省高等学校优秀青年学术带头人支持计划资助
【分类号】：TN249
【正文快照】： 引　言　　国家工信部、财政部联合发布的《煤炭清洁高效利用行动计划(2015—2020)》指出:煤炭是我国的主体能源和重要工业原料,但煤炭利用方式粗放、能效低、污染重等问题没有得到根本解决,需要进一步提高煤炭清洁高效利用水平,强化煤炭利用的能效,确保煤炭利用技术和装置在

【相似文献】