APPLICATION OF EVOLUTIONARY PROGRAMMING TO TRANSIENT AND SUBTRANSIENTPARAMETER-ESTIMATION

This paper presents an artificial intelligence approach of using evolu tionary programming to estimate the transient and subtransient paramet ers of a generator under normal operation, The estimation using evolut ionary programming is compared with that using corrected extended Kalm an filter. The comparisons with both simulation and micromachine test results show that evolutionary programming is robust to search the rea l values of parameters even when the data are highly contaminated by n oises, while with extended Kalman filter, the estimation tends to dive rge with such data.