Nonlinear Observer Design for Power System Monitoring

In this paper, a methodology for designing a nonlinear state observer for power systems is proposed. The algorithm is shown to consist in a sequence of approximate LQ optimal observers whose solution can be explicitely computed thanks to Kalman Filter Riccati equations. A decentralized version of the approach based on augmented Lagrangian relaxation is also discussed. The application to a 5-bus power system is finally presented and demontrates the effectiveness of the approach.