Identification of a Pulverized Fuel Mill From Normal Operating Records Using a Multivariable Polynomial Matrix Model

Abstract Application of a multivariable polynomial matrix model to the structural and parametric identification of a C-type pulverized fuel mill feeding the furnace of a coal-fired power station is presented in this paper. Normal operating records are used and a relation between the boiler master pressure, taken as the output, and primary air differential pressure, mill differential pressure, and coal feeder speed, which were taken as inputs, are identified using the polynomial matrix model. Multi-Input Single Output (MISO) models are fitted using three parameter estimation algorithms and correlation-based model validity checks are applied to test the adequacy of the estimated models. The predictive accuracy of the estimated model is also demonstrated by fitting the model to a second set of data from the same process.