Soft sensing of coal quality

This study presents a soft sensing model of coal quality for utility boilers. This model is based on the coal quality information obtained from exhaust gas. The mechanism modeling method combined with data driving theory is used in the modeling process. The procedure for solving the nonlinear equations for coal quality applies the inner loop iteration of dry ash-free basis of S (Sdaf) and outside loop iteration of dry ash-free basis of N (Ndaf) within a limited range, and dry ash-free basis of C (Cdaf) is searched from the entire range during outside loop iteration. The upper and lower limits of Ndaf are defined according to the NOX content in the exhaust gas, thereby solving the iterative initial value selection problem. Finally, the effectiveness of the proposed method is verified via several simulations and comparisons, the results show that this method is credible and effective and it can be used in power plant for control system optimization.