Indirect identification and control of a laboratory system

The paper describes the application of indirect identification methods to obtain a state-space representation of the open-loop dynamics of a nonlinear laboratory model turbogenerator, which is operated in closed loop. The model structure and selection of variables are determined on the basis of simulation studies and residual analysis, and a recursive instrumental variables technique is employed for online parameter identification. The identified models were used in the design of optimal controllers for the turbogenerator, which have been tested in the laboratory and shown to be most effective. It has also been established that the process of online identification does not reduce the stability of the system.