A Quasilinearization Based Algorithm for the Identification of Transient and Subtransient Parameters of Synchronous Machines

This paper, a result of EPRI RP 1999-5, presents a powerful identification algorithm and assesses its strength for the determination of the transient and subtransient parameters of the synchronous machine. Quasilinearization, the least-squared-error concept and the transition matrix are used to derive the algorithm. In order for this algorithm to identify parameters, a mathematical structure of the system together with the trajectories of all the input variables and a set of output variables captured during a transient performance of the physical system are required. To assess this algorithm, the rotor speed, field voltage, terminal voltages and currents of a generator for a variety of transient performances are simulated. The simulated trajectories are mixed with various levels and types of noise, and are used as input to a computer program in which the identification algorithm under assessment is incorporated. The mathematical structure selected for the synchronous machine in this program, is one of the most detailed one adopted by many utilities. Values determined for transient and subtransient parameters by the identification algorithm are compared with those employed in the simulation program and hereby the strength of the algorithm is assessed.