Research And Implementation Of State Estimation Method Of Power System

Dynamic State Estimation (DSE) is an important part of Energy Management System (EMS) which supplies the state integrated and reliable data for the system. Especially Dynamic State Estimation is a kind of method that models historical data to estimate the new state data. Some new methods of Machine Learning leading into this area solve the bad generalization performance and get higher precision or lower time cost and space cost. For the reason that the survey data in power system is large and nonlinear with high dimensionalities, KPCA and LLE are added into the estimator to pretreat the survey data. Then a new KMPLM estimator is proposed based on the SVM estimator. The experimental results show that these two estimators get good performance at precision of state estimation, time cost of modeling and prediction.