煤炭工业分析转化元素分析计算式改进与回归模型

Calculation formula improvement and regression model for calculate ultimate analysis from proximate analysis of coal

掌握燃煤机组入炉煤的煤质成分,是锅炉安全经济运行的重要依据。对煤质工业分析转化元素分析的计算进行研究,采集了100组煤质成分数据,分析了传统计算公式的不足,并逐一改进碳、氢和氮的计算公式,建立基于支持向量机的元素分析回归模型。此外,对比传统公式、改进公式和回归的计算效果。与传统公式相比,改进公式计算的碳、氢和氮计算的均方根误差R_(MSE)相对降低了16.05%、83.53%和36.36%;在测试集数据中,回归模型在碳、氮和氧的成分计算中R_(MSE)明显较低,分别为0.893、0.116和0.924,而改进公式所计算的氢元素成分R_(MSE)最低(0.037)。

It is an important basis for the safety and economy of the boiler to master the coal quality composition of the coal.The calculation of transformation between element analysis and coal industry analysis was studied.100 groups of data of coal quality composition were collected,and the shortcomings of traditional calculation formulas were analyzed.The formulas for calculating carbon,hydrogen and nitrogen were improved.And the elemental analysis regression model based on support vector was established.Finally,the results of traditional formula,improved formula and regression were compared.The results show that compared with the traditional formula,the R_(MSE)of carbon,hydrogen and nitrogen calculated by the improved formula is reduced by 16.05%,83.53%and 36.36%.In the test set data,the R_(MSE)of the regression model in the composition calculation of carbon,nitrogen and oxygen was lower,which was 0.893,0.116 and 0.924,while the RMSE of hydrogen element calculated by the improved formula is the lowest(0.037).