On-line Identification of Furnace Parameters for Coal-Fired Boiler Control

Abstract

A new identifier for furnace parameters on control coalfired boilers fuel ratio dominating coal burning speed and slagging factor influencing the heat resistance on heating surfaces was applied to a 1000MWe power station. The identifier is based on the Extended Kalman Filtering on a nonlinear boiler dynamics model including energy and mass conservation laws in continuous time with seeking parameters fluctuated as the Brownian motion. The identifier refers to not only measured gas and water temperature at boilers (observable) but also actual fuel flow of each burner (manipulation) for the guidance to plant operators to cope with parameter fluctuation. Due to numerical stability, the identifier uses both Pade (1,1) approximation for time discretizing and U-D factorization for calculating the covariance matrix.