Identification and validation of turbogenerator models

Abstract

Abstract This paper is concerned with the derivation of linear models of a turbogenerator from test data. Data are obtained using a computer-based data acquisition system, which injects small disturbances to the setpoints of plant controllers, and monitors the responses of the unit. The data are processed by a modified Recursive Instrumental Variables algorithm, to provide models of the plant dynamics over a wide range of operating conditions. These state-space models can be estimated on-line, and are intended to provide a basis for multivariable control system design. Tests were performed on a 271 MW unit in a power station, onload and connected to the grid. Results show that accurate models were obtained, and illustrate the range of validity of the models.