Probabilistic robust damping controller designs for FACTS devices and PSS

Abstract

This thesis attempts to improve the previous probabilistic PSS design method and extend it to robust Flexible AC Transmission System (FACTS) controllers design for increasing power system damping under multi-operating conditions. A versatile plug-in modeling technique (PMT), which can model machines and control equipments to any desired degree of complexity, is employed for system state matrix formation. PMT facilitates the expression of eigenvalue sensitivities corresponding to any system parameters. With the variation of operating conditions regarded as disturbing resources, probabilistic load flow (PLF) is applied firstly for calculating the numerical attributes of nodal voltages. The probabilistic distribution of system eigenvalue is then obtained by a hybrid utilization of central moments and cumulants. Probabilstic sensitivity index (PSI) is also derived, whereby the conventional residue index and participation factor are extended to probabilistic environment. When calculating PLF and probabilistic eigenvalue, several models considering different approximation degree of second order term are derived and their precisions are compared with the actual results by repeatedly deterministic method, in order to select the most suitable model. For calculating probabilistic eigenvalue, the random variables can be of any distributions. But for probabilistic eigenvalue sensitivity analysis for damping controller design, normal distribution has to be assumed because of the difficulty to gain the probabilistic eigenvalue sensitivity index (PSI) under arbitrary distribution. Consequently, a systematic probabilistic analysis is proposed for PSS and FACTS controllers' suitable location/input signals selection and parameters tuning. As for PSS, the damping torque component contributed by PSS is examined by modal analysis. As for FACTS controllers (thyristor-controlled series capacitor (TCSC) and static-var compensator (SVC)), two test systems are adopted for verifying SVC and TCSC design. To improve the system damping with multi controllers, the coordination between PSS and FACTS device is proposed by an optimization-based tuning algorithm, taking into account the interactions among PSS, SVC and TCSC. By minimizing the objective function, the overall system performance is much enhanced.