摘要
脱硝效率受到众多关联性较强的因素影响,使得脱硝效率难以准确地实时测量。采用皮尔逊相关系数与反应机理结合的方法选取辅助变量,运用支持向量机建立以选取的辅助变量为输入脱硝效率为输出的预测模型。并基于某电厂脱硝系统实际运行数据对模型进行训练验证。结果表明:预测模型相关系数高达99.897 9%,均方误差为6.574 68×10^(-5);大部分的样本点误差在-1.0~1.0之间,部分误差趋于0值,最大误差绝对值不超过1.8。说明该模型的预测精度较高,且具有良好的推广能力,能较好地满足工程实际需求。
The denitrification efficiency is influenced by many factors with strong correlation,which makes the denitrification efficiency difficult to measure in real time. In this paper,the auxiliary variables are selected by the combination of Pearson correlation coefficient and response mechanism,and the support vector machine is used to establish the prediction model with auxiliary variables as input and denitrification efficiency as output. And then the model is trained and verified based on the actual operation data of the denitrification system of a power plant. The results show that the correlation coefficient is 99. 8979% and the mean square error is 6. 57468 × 10-5. Most of the sample points have the errors between-1. 0 and 1. 0,and some errors tend to 0,and the maximum error is no more than 1. 8. It shows that the prediction accuracy of the model is good and satisfactory,and it has good generalization ability and can meet the practical needs.
出处
《电力科学与工程》
2017年第10期34-39,共6页
Electric Power Science and Engineering
基金
国家自然科学基金(51606066)
关键词
脱硝效率
皮尔逊相关系数
反应机理
支持向量机
预测模型
denitrification efficiency
Pearson correlation coefficient
response mechanism
support vector machine
prediction model