A DYNAMIC ONLINE PARAMETER-IDENTIFICATION AND FULL-SCALE SYSTEM EXPERIMENTAL-VERIFICATION FOR LARGE SYNCHRONOUS MACHINES

An on-line parameter identification and full-scale experimental verifi cation for large synchronous machines (>50 MVA) is presented in this p aper. A step change of excitation is imposed to a generator when the m achine is in normal operation. The transient voltages, currents and th e power angle are recorded. Based on the large disturbance equations a nd using the measured power angle as an observation argument in an ide ntification algorithm, the synchronous machine electrical parameters ( x(d), x(d)', x(d) '', T-do', T-do'', x(q), x(q) '', T-qo'') and mechan ical parameters (H, D) are obtained. In addition, the system parameter s (equivalent infinite bus voltage V-bus and line reactance x(e)) are identified as well. The proposed method has been repetitively applied to turbogenerators and hydrogenerators with capacities up to 300 MVA. In particular, a field test has been conducted on a system with a capa city of 15,000 MVA. The experimental results from all of the full-scal e tests are consistent and the effectiveness of the proposed on-line i dentification method is verified. The plant experiences indicate that by adopting the identified parameters, the stability margin of the gen erator can be improved up to 5%, resulting in 30-50 MVA more power gen eration.