Optimal Design of a PMU-Based Monitoring Architecture for Power Systems

In this paper, a methodology for designing an optimal monitoring architecture by using a limited number of PMU (Phasor Measurement Units) and a PDC (Phase Data Concentrator) is proposed. The optimal design problem consists in defining the optimal location of both the PMU and the PDC by maximizing the expected value of the trace of the observability gramian of the power system over a large number of set point scenarii, while minimizing the communication infrastructure cost. Furthermore, a nonlinear dynamical EKF (Extended Kalman Filtering) state-observer is proposed. This state-observer allows to take transient phenomena into account for wide-area power systems described by algebraic-differential equations, without needing nonlinear inversion techniques. Application to a 5-bus power system is also presented and demontrates the effectiveness of the approach.