Dynamic prediction model for NOx emission at the outlet of SCR system based on extreme learning machine

Selective catalytic reduction (SCR) system with high efficiency has been widely applied for reducing Nitrogen oxide (NOx) emissions in thermal power plants. Establishing an accurate dynamic prediction model is important for optimal control of SCR system. Firstly, data preprocessing is conducted maximal information coefficient (MIC) is used to estimate the including processing outlets and normalization; Secondly, the delay time of each variable in order to reconstruct the data. Then, develop a new feature selection method by using Lasso combined with partial least squares (PLS) regression; Finally, the dynamic prediction model is established based on extreme learning machine (ELM) with the input variables selected before. Based on the actual historical operating data, the experimental results show that the model established in this study can accurately predict the NOx emission at the outlet of SCR system.