Nonlinear Estimation of Salient-Pole Synchronous Machines Parameters via Particle Filter

Some of the major challenges in Smart Grids protection and control studies are related to system stability and modeling issues. Synchronous machines are usually present in those systems as generators and, thus, play a key role in these matters. This work presents a non-linear and real-time approach to estimate direct- and quadrature-axis magnetizing reactances and states of synchronous machines, such as fluxes linkages, rotor speed, and power angle, by means of particle filtering. An eighth order model – which represents the electromechanical behavior during transient- and steady-state – is considered and only measurements taken at the machine armature and field terminals during steady-state are necessary to estimate the states and parameters. The simulated results presented validate the accuracy of the proposed method.