垃圾焚烧烟气中二噁英类浓度的支持向量回归预测  

Prediction of dioxin emissions in flue gas from waste incineration based on support vector regression

作　　者：肖晓东[1,2] 卢加伟[2] 海景[2] 廖利[1] 

机构地区：[1]华中科技大学环境科学与工程学院,湖北武汉430074 [2]环境保护部华南环境科学研究所,广东广州510655

出　　处：《可再生能源》2017年第8期1107-1114,共8页Renewable Energy Resources

基　　金：国家自然科学基金项目(51608223);中央级公益性科研院所基本科研业务专项(PM-zx703-201602-050)

摘　　要：文章利用垃圾焚烧烟气的在线监测数据,探索建立二噁英类浓度与烟气中其他污染物浓度、工况参数的多元回归预测模型,从而实现对烟气中二噁英类浓度的间接实时监测,但受限于二噁英类浓度的监测成本高,可供回归预测训练的样本少的问题,线性回归预测模型的泛化能力较弱、稳定性较差。使用非线性的支持向量回归方法建模,有助于解决这一问题。文章使用3种核函数构建支持向量回归预测模型,应用华南地区某垃圾焚烧厂的10组监测数据作为训练集和测试集,比较了支持向量回归预测和多元线性回归预测的相对误差。研究结果表明,训练集为8组数据时,支持向量回归预测的相对误差明显小于多元线性回归,尤其是模型使用1阶多项式核函数和径向基核函数时的最大百分比误差较小,泛化能力较强。It is difficult to monitor dioxin emissions in flue gas from waste incineration in real time because of the high cost. However, multiple regressions can be applied to predicting dioxin emissions in flue gas, with the aid of pollutant concentrations and operating parameters, which are collected by on-line monitoring systems. The critical problem is that there are very few samples of dioxin available for training prediction models due to the high cost. This leads to a poor performance of generalization when using linear regression models. A nonlinear method named support vector regression is presented in this paper, in order to improve the generalization performance of prediction. This paper compares support vector regression models using three different kernel functions with the multiple linear regression model. Relative errors are calculated to evaluate the generalization performance with the aid of 10 sets of data monitored on a waste incineration plant in South China. The result shows support vector regression models have lower relative errors. The maximum possible percentage errors of models using first-order polynomial kernel function and radial basis kernel function are much lower than the model using Sigmoid kernel function, especially.

关 键 词：垃圾焚烧 烟气 二噁英类 小样本 支持向量回归 

分 类 号：TK6[动力工程及工程热物理—生物能] X705[环境科学与工程—环境工程]