An adaptive State estimator for pulverizer control using moments of particle size distribution

An adaptive state estimator for pulverizers consisting of blending, grinding, and classifying processes has been developed in order to improve control of pulverized-coal-fired power stations. Though coal flow and non-Gaussian particle size distributions in the processes are mutually related, the estimator is able to efficiently simulate flow and normalized moments of the distributions with a state vector. The estimator also identifies coal grindability for adapting to variation in coal characteristic in parallel with the process simulation. The accuracy of the adaptive estimation and the effectiveness in improving the load-swinging performance have been validated at a 1000-MWe class power station.