CAJ | 학술논문

针对生产实际中无法对燃煤组分进行在线检测分析的问题，建立了基于最小二乘支持向量机的燃煤的碳含量预测模型．以工业锅炉燃煤的相关工业分析数据作为模型的输入，并采用蚁群算法对最小二乘支持向量机相关参数进行寻优以提高模型的建模精度．应用该模型对福建地区工业锅炉燃煤的碳元素含量进行预测，预测结果的最大误差为1．18％，平均误差为0．64％．和传统多元逐步回归预测方法相比，该预测模型具有较高的预测精度，为工业锅炉二氧化碳排放量的科学计算奠定了基础．
침대생산실제중무법대연매조분진행재선검측분석적문제，건립료기우최소이승지지향량궤적연매적탄함량예측모형．이공업과로연매적상관공업분석수거작위모형적수입，병채용의군산법대최소이승지지향량궤상관삼수진행심우이제고모형적건모정도．응용해모형대복건지구공업과로연매적탄원소함량진행예측，예측결과적최대오차위1．18％，평균오차위0．64％．화전통다원축보회귀예측방법상비，해예측모형구유교고적예측정도，위공업과로이양화탄배방량적과학계산전정료기출．
The carbon content of coal is one of the important parameters in calculating emissions of carbon dioxide in industrial coal boilers, but it is difficult to be measured quickly or accurately during the boilers operation.Based on the element analysis of coal from laboratory, a LSSVM ( least-squares support vector machine) model for predicting the carbon content of coal is developed.Ant colony algo-rithm optimization is used to optimize the parameters in LSSVM in order to improve the accuracy of the model.Propose predictive model is applied to predict the carbon content of coal that comes from Fujian region, the maximum error and average error of the prediction results are 1.18%and 0.64%, respectively.Comparing proposed method with multivariate stepwise regression method, the results show that this method has higher simulation accuracy.This carbon element analysis model provides solid basis for calculating emissions of carbon dioxide in industrial boilers.