Microgrids have become a feasible solution to the problems being faced in the conventional grid. Real time monitoring and control of microgrid can be achieved by using Dynamic State Estimation (DSE) which keeps track of the dynamic state variables. DSE models are solved by using non-linear variants of Kalman filters such as Extended Kalman Filter (EKF), Unscented Kalman Filter (UKF), Square Root UKF (SRUKF) etc. In EKF, the non-linear function is linearized by using first order Taylor series expansion, which provides good estimates with less non- linearity. During high non linearity situation the EKF gives biased estimates for transmission network. The DSE formulation depends on the admittance matrix of the network, which needs to be re-computed for changes in the network configuration, making it less reliable.



This seminar talk proposes a new formulation of DSE which is completely independent of the admittance matrix of the network. A Robust Unscented Kalman filter technique is proposed for DSE of microgrid for improving survivability of DERs in order to prevent them from stalling. Detailed Dynamic Modelling of the network, synchronous generator DER and micro grid will be presented. The working of the method is implemented for a CERTS microgrid and in a 13.8 kV distribution network which can operate either in grid connected or an islanded mode. The results of the scenarios of EKF, UKF, SRUKF, RUKF are presented with improved computation time. In UKF there are chances that error covariance may lose its positive definiteness property which eventually leads to failure of the filter. This problem can be overcome by using SRUKF. RUKF and SRUKF gives accurate estimates of the states when compared with other techniques. However, the computational time for RUKF is less than SRUKF. Hence the proposed algorithm may be used in real time applications for DSE in a microgrid.