Advanced network modelling techniques for power system state estimation

Abstract

State estimation is the process to calculate estimates for unknown state variables of a power network by minimising an objective function. Classical state estimators take into account that network parameters e.g. impedances and ad­mittances, and topology of the power network are accurately known for state estimation. This assumption can cause state variables to be incorrectly estimated. Network impedances can be in­correct or even unknown when a state estimation process needs to be done. The topology of the power network can also be incorrect or even unknown. A solution for these problems is to use a gen­eralised state estimator. A generalised state estimator uses advanced network modelling techniques. These techniques can be used to give estimates for network parameters and topology. The following advanced network modelling techniques are introduced: • Modelling of circuit breakers • Modelling of zero impedance branches • Modelling of network parameters • Modelling of transformer parameters These advanced modelling techniques are used in a state estimation process which minimised a Weighted Least Squares objective function. These advanced modelling techniques are also used in a state estimation process which minimised a Quadratic-Tangent objective function. Measurements obtained from a network are fraught with errors/bad data. Bad data is used in a state estimation process which minimised a Weighted Least Squares objective function. Bad data is also used in a state estimation process which minimised a Quadratic-Tangent objective function. Advanced network modelling techniques are used in both these estimation processes. From the results obtained, it is shown that in the presence of bad data, advanced network modelling techniques can be used in a state estimation process which minimises a Quadratic-Tangent objective function.