Effective dynamic state estimation algorithm for Islanded microgrid structures based on singular perturbation theory
Por:
Vieyra, Natanael, Maya, Paul, Castro, Luis M.
Publicada:
1 oct 2020
Resumen:
This paper introduces an effective dynamic state estimator for Islanded
Microgrids. Basing on a set of nonlinear Differential Algebraic
equations representing the electrical grid and the energy sources, the
Singular Perturbation Theory is used to obtain a modified mathematical
representation of the Microgrid model to develop an effective dynamic
state estimator based on the Unscented Kalman Filter. It is shown that
Singular Perturbation Theory is a viable tool that permits the design of
a dynamic estimator able to effectively recover the steady-state and
dynamic states of the electrical grid, that is, the nodal voltages and
the dynamic variables of generator units. Furthermore, the Microgrid
state is suitably recovered using fewer measurements than those needed
by conventional static estimators. The performance of the proposed
scheme is evaluated using a practical Microgrid containing wind power
and hydroelectric generators, under load and wind variations as well as
three-phase faults. Also, this timely approach is compared with the
Extended Kalman Filter for Differential Algebraic systems, demonstrating
the superior effectiveness of the developed state estimator: the errors
obtained by the new dynamic state estimator are 80% smaller than those
obtained by the conventional Extended Kalman Filter, for the same
applied noises. Moreover, a comparative study case with the Unscented
Kalman Filter is included.
Filiaciones:
Vieyra, Natanael:
Univ Nacl Autonoma Mexico, Dept Control & Robot, Mexico City, DF, Mexico
Maya, Paul:
Univ Nacl Autonoma Mexico, Dept Control & Robot, Mexico City, DF, Mexico
Castro, Luis M.:
Univ Nacl Autonoma Mexico, Dept Elect Energy, Mexico City, DF, Mexico
|