Modeling of flue gas path system for desulfurization in power plant with k-fold cross validation

Abstract

© 2014 IEEE. A mathematical system model is usually construct to capture dominant dynamical behaviors of an expensive plant for analysis. This approach is undertaken to minimize unforeseen damages and operation interruptions from experimenting on the actual system plant. In this paper, a procedure of mathematical modeling of flue gas path system for wet desulfurization with lime in Mae Moh coal power plant, Thailand is presented. The modeling is based on the input-output of subsystems with interconnections similar to the engine-breathing turbocharged diesel engine model. Specifically, this model is derived based on the mass conservation law and the ideal gas assumption. A difference equation is derived for a simulation evaluation by applying Forward Euler rule. Due to the limitation on the quality of the collected data, linear interpolation with delay technique is applied as data treatment. Then the model parameters are computed in the estimating process. The model parameters are validated against the collected actual data assessing the performance of the proposed system model. System model parameters are identified based on a least squares technique in a k-fold cross validation. The system model performance and errors are given in terms of root-mean-square error (RMSE). Based on the simulation results, it is found that the proposed system model is promising and delivers satisfactory performance under normal load changing conditions.