Contents I. Introduction
The power system security needs to have a detailed monitor to the operating conditions of the system. In the control center, the state estimator can deal with the measurements received from the remote terminal units (RTUs) at the substations. Those measurements include bus voltages, branch currents, real and reactive power flows, and power injections. Recently, phasor measurement units (PMUs) with the help of global positioning system (GPS) can provide synchronized phasor measurements of voltages and currents [1]. That will improve the performance of the state estimator because the synchronized measurements provided by PMUs during the dynamic event will make the state estimation more accurate [2]. When a PMU is placed at one bus, it can measure the bus phasor voltage and phasor currents of all lines connected to that bus making the system observable [3]. However, integrating PMUs at each bus instantaneously is difficult and expensive. Then the problem that needs to be solved is to find the optimal PMU placement that can make the system observable using PMUs at certain buses. In this paper, the optimal PMU placement (OPP) problem is solved using two different approaches which are mixed integer linear programming (MILP) and nonlinear programming (NLP). The system is then studied over several cases including power flow measurements, zero injection, limited communication facility, and single PMU failure. A comparison between the two approaches is conducted to show their advantages and disadvantages. In addition, system observability redundancy index (SORI) is demonstrated to provide a higher redundancy. Therefore, the objective of this formulation is to minimize the total number of PMUs and maximize the redundancy measurements in power systems.
